Release 1.4.8

- Do not use \r if output is not a terminal (should fix unexpected job output in proxmox)
- Fix rm/rm-data error return code, add --down-ok option to bypass the error
- Add EIO retry timeout and allow to disable these retries, rename up_wait_retry_interval to client_retry_interval
- Add ubuntu jammy build
- Wait for blockstore initialisation before starting OSD (prevent timeouts when init takes time)
- Fix a rare use-after-free in automatic sync after delete in blockstore

.gitea/workflows/test.yml

@@ -395,7 +395,7 @@ jobs:
     steps:
     - name: Run test
       id: test
-      timeout-minutes: 3
+      timeout-minutes: 6
       run: SCHEME=ec /root/vitastor/tests/test_snapshot_chain.sh
     - name: Print logs
       if: always() && steps.test.outcome == 'failure'

.gitea/workflows/tests-to-yaml.pl

@@ -39,6 +39,10 @@ for my $line (<>)
                 $test_name .= '_'.lc($1).'_'.$2;
             }
         }
+        if ($test_name eq 'test_snapshot_chain_ec')
+        {
+            $timeout = 6;
+        }
         $line =~ s!\./test_!/root/vitastor/tests/test_!;
         # Gitea CI doesn't support artifacts yet, lol
         #- name: Upload results

CMakeLists.txt

@@ -2,6 +2,6 @@ cmake_minimum_required(VERSION 2.8.12)
 
 project(vitastor)
 
-set(VERSION "1.4.2")
+set(VERSION "1.4.8")
 
 add_subdirectory(src)

csi/Makefile

@@ -1,4 +1,4 @@
-VERSION ?= v1.4.2
+VERSION ?= v1.4.8
 
 all: build push
 

csi/deploy/004-csi-nodeplugin.yaml

@@ -49,7 +49,7 @@ spec:
             capabilities:
               add: ["SYS_ADMIN"]
             allowPrivilegeEscalation: true
-          image: vitalif/vitastor-csi:v1.4.2
+          image: vitalif/vitastor-csi:v1.4.8
           args:
             - "--node=$(NODE_ID)"
             - "--endpoint=$(CSI_ENDPOINT)"

csi/deploy/007-csi-provisioner.yaml

@@ -121,7 +121,7 @@ spec:
             privileged: true
             capabilities:
               add: ["SYS_ADMIN"]
-          image: vitalif/vitastor-csi:v1.4.2
+          image: vitalif/vitastor-csi:v1.4.8
           args:
             - "--node=$(NODE_ID)"
             - "--endpoint=$(CSI_ENDPOINT)"

csi/src/config.go

@@ -5,7 +5,7 @@ package vitastor
 
 const (
     vitastorCSIDriverName    = "csi.vitastor.io"
-    vitastorCSIDriverVersion = "1.4.2"
+    vitastorCSIDriverVersion = "1.4.8"
 )
 
 // Config struct fills the parameters of request or user input

debian/build-vitastor-bookworm.sh

@@ -3,5 +3,5 @@
 cat < vitastor.Dockerfile > ../Dockerfile
 cd ..
 mkdir -p packages
-sudo podman build --build-arg REL=bookworm -v `pwd`/packages:/root/packages -f Dockerfile .
+sudo podman build --build-arg DISTRO=debian --build-arg REL=bookworm -v `pwd`/packages:/root/packages -f Dockerfile .
 rm Dockerfile

debian/build-vitastor-bullseye.sh

@@ -3,5 +3,5 @@
 cat < vitastor.Dockerfile > ../Dockerfile
 cd ..
 mkdir -p packages
-sudo podman build --build-arg REL=bullseye -v `pwd`/packages:/root/packages -f Dockerfile .
+sudo podman build --build-arg DISTRO=debian --build-arg REL=bullseye -v `pwd`/packages:/root/packages -f Dockerfile .
 rm Dockerfile

debian/build-vitastor-buster.sh

@@ -3,5 +3,5 @@
 cat < vitastor.Dockerfile > ../Dockerfile
 cd ..
 mkdir -p packages
-sudo podman build --build-arg REL=buster -v `pwd`/packages:/root/packages -f Dockerfile .
+sudo podman build --build-arg DISTRO=debian --build-arg REL=buster -v `pwd`/packages:/root/packages -f Dockerfile .
 rm Dockerfile

debian/build-vitastor-ubuntu-jammy.sh

@@ -0,0 +1,7 @@
+#!/bin/bash
+
+cat < vitastor.Dockerfile > ../Dockerfile
+cd ..
+mkdir -p packages
+sudo podman build --build-arg DISTRO=ubuntu --build-arg REL=jammy -v `pwd`/packages:/root/packages -f Dockerfile .
+rm Dockerfile

debian/changelog

@@ -1,4 +1,4 @@
-vitastor (1.4.2-1) unstable; urgency=medium
+vitastor (1.4.8-1) unstable; urgency=medium
 
   * Bugfixes
 

debian/libvirt.Dockerfile

@@ -1,13 +1,14 @@
 # Build patched libvirt for Debian Buster or Bullseye/Sid inside a container
-# cd ..; podman build --build-arg REL=bullseye -v `pwd`/packages:/root/packages -f debian/libvirt.Dockerfile .
+# cd ..; podman build --build-arg DISTRO=debian --build-arg REL=bullseye -v `pwd`/packages:/root/packages -f debian/libvirt.Dockerfile .
 
+ARG DISTRO=
 ARG REL=
-FROM debian:$REL
+FROM $DISTRO:$REL
 ARG REL=
 
 WORKDIR /root
 
-RUN if [ "$REL" = "buster" -o "$REL" = "bullseye" ]; then \
+RUN if ([ "${DISTRO}" = "debian" ]) && ( [ "${REL}" = "buster" -o "${REL}" = "bullseye" ] ); then \
         echo "deb http://deb.debian.org/debian $REL-backports main" >> /etc/apt/sources.list; \
         echo >> /etc/apt/preferences; \
         echo 'Package: *' >> /etc/apt/preferences; \
@@ -23,7 +24,7 @@ RUN apt-get -y build-dep libvirt0
 RUN apt-get -y install libglusterfs-dev
 RUN apt-get --download-only source libvirt
 
-ADD patches/libvirt-5.0-vitastor.diff patches/libvirt-7.0-vitastor.diff patches/libvirt-7.5-vitastor.diff patches/libvirt-7.6-vitastor.diff /root
+ADD patches/libvirt-5.0-vitastor.diff patches/libvirt-7.0-vitastor.diff patches/libvirt-7.5-vitastor.diff patches/libvirt-7.6-vitastor.diff patches/libvirt-8.0-vitastor.diff /root
 RUN set -e; \
     mkdir -p /root/packages/libvirt-$REL; \
     rm -rf /root/packages/libvirt-$REL/*; \

debian/vitastor.Dockerfile

@@ -1,8 +1,10 @@
 # Build Vitastor packages for Debian inside a container
-# cd ..; podman build --build-arg REL=bullseye -v `pwd`/packages:/root/packages -f debian/vitastor.Dockerfile .
+# cd ..; podman build --build-arg DISTRO=debian --build-arg REL=bullseye -v `pwd`/packages:/root/packages -f debian/vitastor.Dockerfile .
 
+ARG DISTRO=debian
 ARG REL=
-FROM debian:$REL
+FROM $DISTRO:$REL
+ARG DISTRO=debian
 ARG REL=
 
 WORKDIR /root
@@ -35,8 +37,8 @@ RUN set -e -x; \
     mkdir -p /root/packages/vitastor-$REL; \
     rm -rf /root/packages/vitastor-$REL/*; \
     cd /root/packages/vitastor-$REL; \
-    cp -r /root/vitastor vitastor-1.4.2; \
+    cp -r /root/vitastor vitastor-1.4.8; \
-    cd vitastor-1.4.2; \
+    cd vitastor-1.4.8; \
     ln -s /root/fio-build/fio-*/ ./fio; \
     FIO=$(head -n1 fio/debian/changelog | perl -pe 's/^.*\((.*?)\).*$/$1/'); \
     ls /usr/include/linux/raw.h || cp ./debian/raw.h /usr/include/linux/raw.h; \
@@ -49,8 +51,8 @@ RUN set -e -x; \
     rm -rf a b; \
     echo "dep:fio=$FIO" > debian/fio_version; \
     cd /root/packages/vitastor-$REL; \
-    tar --sort=name --mtime='2020-01-01' --owner=0 --group=0 --exclude=debian -cJf vitastor_1.4.2.orig.tar.xz vitastor-1.4.2; \
+    tar --sort=name --mtime='2020-01-01' --owner=0 --group=0 --exclude=debian -cJf vitastor_1.4.8.orig.tar.xz vitastor-1.4.8; \
-    cd vitastor-1.4.2; \
+    cd vitastor-1.4.8; \
     V=$(head -n1 debian/changelog | perl -pe 's/^.*\((.*?)\).*$/$1/'); \
     DEBFULLNAME="Vitaliy Filippov <vitalif@yourcmc.ru>" dch -D $REL -v "$V""$REL" "Rebuild for $REL"; \
     DEB_BUILD_OPTIONS=nocheck dpkg-buildpackage --jobs=auto -sa; \

docs/config/client.en.md

@@ -9,6 +9,8 @@
 These parameters apply only to Vitastor clients (QEMU, fio, NBD and so on) and
 affect their interaction with the cluster.
 
+- [client_retry_interval](#client_retry_interval)
+- [client_eio_retry_interval](#client_eio_retry_interval)
 - [client_max_dirty_bytes](#client_max_dirty_bytes)
 - [client_max_dirty_ops](#client_max_dirty_ops)
 - [client_enable_writeback](#client_enable_writeback)
@@ -19,6 +21,26 @@ affect their interaction with the cluster.
 - [nbd_max_devices](#nbd_max_devices)
 - [nbd_max_part](#nbd_max_part)
 
+## client_retry_interval
+
+- Type: milliseconds
+- Default: 50
+- Minimum: 10
+- Can be changed online: yes
+
+Retry time for I/O requests failed due to inactive PGs or network
+connectivity errors.
+
+## client_eio_retry_interval
+
+- Type: milliseconds
+- Default: 1000
+- Can be changed online: yes
+
+Retry time for I/O requests failed due to data corruption or unfinished
+EC object deletions (has_incomplete PG state). 0 disables such retries
+and clients are not blocked and just get EIO error code instead.
+
 ## client_max_dirty_bytes
 
 - Type: integer

docs/config/client.ru.md

@@ -9,6 +9,8 @@
 Данные параметры применяются только к клиентам Vitastor (QEMU, fio, NBD и т.п.) и
 затрагивают логику их работы с кластером.
 
+- [client_retry_interval](#client_retry_interval)
+- [client_eio_retry_interval](#client_eio_retry_interval)
 - [client_max_dirty_bytes](#client_max_dirty_bytes)
 - [client_max_dirty_ops](#client_max_dirty_ops)
 - [client_enable_writeback](#client_enable_writeback)
@@ -19,6 +21,27 @@
 - [nbd_max_devices](#nbd_max_devices)
 - [nbd_max_part](#nbd_max_part)
 
+## client_retry_interval
+
+- Тип: миллисекунды
+- Значение по умолчанию: 50
+- Минимальное значение: 10
+- Можно менять на лету: да
+
+Время повтора запросов ввода-вывода, неудачных из-за неактивных PG или
+ошибок сети.
+
+## client_eio_retry_interval
+
+- Тип: миллисекунды
+- Значение по умолчанию: 1000
+- Можно менять на лету: да
+
+Время повтора запросов ввода-вывода, неудачных из-за повреждения данных
+или незавершённых удалений EC-объектов (состояния PG has_incomplete).
+0 отключает повторы таких запросов и клиенты не блокируются, а вместо
+этого просто получают код ошибки EIO.
+
 ## client_max_dirty_bytes
 
 - Тип: целое число

docs/config/monitor.en.md

@@ -19,8 +19,8 @@ These parameters only apply to Monitors.
 ## etcd_mon_ttl
 
 - Type: seconds
-- Default: 30
+- Default: 1
-- Minimum: 10
+- Minimum: 5
 
 Monitor etcd lease refresh interval in seconds
 

docs/config/monitor.ru.md

@@ -19,8 +19,8 @@
 ## etcd_mon_ttl
 
 - Тип: секунды
-- Значение по умолчанию: 30
+- Значение по умолчанию: 1
-- Минимальное значение: 10
+- Минимальное значение: 5
 
 Интервал обновления etcd резервации (lease) монитором
 

docs/config/network.en.md

@@ -25,7 +25,6 @@ between clients, OSDs and etcd.
 - [peer_connect_timeout](#peer_connect_timeout)
 - [osd_idle_timeout](#osd_idle_timeout)
 - [osd_ping_timeout](#osd_ping_timeout)
-- [up_wait_retry_interval](#up_wait_retry_interval)
 - [max_etcd_attempts](#max_etcd_attempts)
 - [etcd_quick_timeout](#etcd_quick_timeout)
 - [etcd_slow_timeout](#etcd_slow_timeout)
@@ -212,17 +211,6 @@ Maximum time to wait for OSD keepalive responses. If an OSD doesn't respond
 within this time, the connection to it is dropped and a reconnection attempt
 is scheduled.
 
-## up_wait_retry_interval
-
-- Type: milliseconds
-- Default: 500
-- Minimum: 50
-- Can be changed online: yes
-
-OSDs respond to clients with a special error code when they receive I/O
-requests for a PG that's not synchronized and started. This parameter sets
-the time for the clients to wait before re-attempting such I/O requests.
-
 ## max_etcd_attempts
 
 - Type: integer

docs/config/network.ru.md

@@ -25,7 +25,6 @@
 - [peer_connect_timeout](#peer_connect_timeout)
 - [osd_idle_timeout](#osd_idle_timeout)
 - [osd_ping_timeout](#osd_ping_timeout)
-- [up_wait_retry_interval](#up_wait_retry_interval)
 - [max_etcd_attempts](#max_etcd_attempts)
 - [etcd_quick_timeout](#etcd_quick_timeout)
 - [etcd_slow_timeout](#etcd_slow_timeout)
@@ -221,19 +220,6 @@ OSD в любом случае согласовывают реальное зн
 Если OSD не отвечает за это время, соединение отключается и производится
 повторная попытка соединения.
 
-## up_wait_retry_interval
-
-- Тип: миллисекунды
-- Значение по умолчанию: 500
-- Минимальное значение: 50
-- Можно менять на лету: да
-
-Когда OSD получают от клиентов запросы ввода-вывода, относящиеся к не
-поднятым на данный момент на них PG, либо к PG в процессе синхронизации,
-они отвечают клиентам специальным кодом ошибки, означающим, что клиент
-должен некоторое время подождать перед повторением запроса. Именно это время
-ожидания задаёт данный параметр.
-
 ## max_etcd_attempts
 
 - Тип: целое число

docs/config/osd.en.md

@@ -59,6 +59,7 @@ them, even without restarting by updating configuration in etcd.
 - [recovery_tune_client_util_high](#recovery_tune_client_util_high)
 - [recovery_tune_agg_interval](#recovery_tune_agg_interval)
 - [recovery_tune_sleep_min_us](#recovery_tune_sleep_min_us)
+- [recovery_tune_sleep_cutoff_us](#recovery_tune_sleep_cutoff_us)
 
 ## etcd_report_interval
 
@@ -604,5 +605,14 @@ is usually fine.
 - Default: 10
 - Can be changed online: yes
 
-Minimum possible value for auto-tuned recovery_sleep_us. Values lower
+Minimum possible value for auto-tuned recovery_sleep_us. Lower values
-than this value are changed to 0.
+are changed to 0.
+
+## recovery_tune_sleep_cutoff_us
+
+- Type: microseconds
+- Default: 10000000
+- Can be changed online: yes
+
+Maximum possible value for auto-tuned recovery_sleep_us. Higher values
+are treated as outliers and ignored in aggregation.

docs/config/osd.ru.md

@@ -60,6 +60,7 @@
 - [recovery_tune_client_util_high](#recovery_tune_client_util_high)
 - [recovery_tune_agg_interval](#recovery_tune_agg_interval)
 - [recovery_tune_sleep_min_us](#recovery_tune_sleep_min_us)
+- [recovery_tune_sleep_cutoff_us](#recovery_tune_sleep_cutoff_us)
 
 ## etcd_report_interval
 
@@ -634,4 +635,14 @@ EC (кодов коррекции ошибок) с более, чем 1 диск
 - Можно менять на лету: да
 
 Минимальное возможное значение авто-подстроенного recovery_sleep_us.
-Значения ниже данного заменяются на 0.
+Меньшие значения заменяются на 0.
+
+## recovery_tune_sleep_cutoff_us
+
+- Тип: микросекунды
+- Значение по умолчанию: 10000000
+- Можно менять на лету: да
+
+Максимальное возможное значение авто-подстроенного recovery_sleep_us.
+Большие значения считаются случайными выбросами и игнорируются в
+усреднении.

docs/config/pool.en.md

@@ -154,6 +154,9 @@ That is, if it becomes impossible to place PG data on at least (pg_minsize)
 OSDs, PG is deactivated for both read and write. So you know that a fresh
 write always goes to at least (pg_minsize) OSDs (disks).
 
+That is, pg_size minus pg_minsize sets the number of disk failures to tolerate
+without temporary downtime (for [osd_out_time](monitor.en.md#osd_out_time)).
+
 FIXME: pg_minsize behaviour may be changed in the future to only make PGs
 read-only instead of deactivating them.
 

docs/config/pool.ru.md

@@ -157,6 +157,10 @@
 OSD, PG деактивируется на чтение и запись. Иными словами, всегда известно,
 что новые блоки данных всегда записываются как минимум на pg_minsize дисков.
 
+По сути, разница pg_size и pg_minsize задаёт число отказов дисков, которые пул
+может пережить без временной (на [osd_out_time](monitor.ru.md#osd_out_time))
+остановки обслуживания.
+
 FIXME: Поведение pg_minsize может быть изменено в будущем с полной деактивации
 PG на перевод их в режим только для чтения.
 

docs/config/src/client.yml

@@ -1,3 +1,27 @@
+- name: client_retry_interval
+  type: ms
+  min: 10
+  default: 50
+  online: true
+  info: |
+    Retry time for I/O requests failed due to inactive PGs or network
+    connectivity errors.
+  info_ru: |
+    Время повтора запросов ввода-вывода, неудачных из-за неактивных PG или
+    ошибок сети.
+- name: client_eio_retry_interval
+  type: ms
+  default: 1000
+  online: true
+  info: |
+    Retry time for I/O requests failed due to data corruption or unfinished
+    EC object deletions (has_incomplete PG state). 0 disables such retries
+    and clients are not blocked and just get EIO error code instead.
+  info_ru: |
+    Время повтора запросов ввода-вывода, неудачных из-за повреждения данных
+    или незавершённых удалений EC-объектов (состояния PG has_incomplete).
+    0 отключает повторы таких запросов и клиенты не блокируются, а вместо
+    этого просто получают код ошибки EIO.
 - name: client_max_dirty_bytes
   type: int
   default: 33554432

docs/config/src/network.yml

@@ -243,21 +243,6 @@
     Максимальное время ожидания ответа на запрос проверки состояния соединения.
     Если OSD не отвечает за это время, соединение отключается и производится
     повторная попытка соединения.
-- name: up_wait_retry_interval
-  type: ms
-  min: 10
-  default: 50
-  online: true
-  info: |
-    OSDs respond to clients with a special error code when they receive I/O
-    requests for a PG that's not synchronized and started. This parameter sets
-    the time for the clients to wait before re-attempting such I/O requests.
-  info_ru: |
-    Когда OSD получают от клиентов запросы ввода-вывода, относящиеся к не
-    поднятым на данный момент на них PG, либо к PG в процессе синхронизации,
-    они отвечают клиентам специальным кодом ошибки, означающим, что клиент
-    должен некоторое время подождать перед повторением запроса. Именно это время
-    ожидания задаёт данный параметр.
 - name: max_etcd_attempts
   type: int
   default: 5

docs/config/src/osd.yml

@@ -731,8 +731,19 @@
   default: 10
   online: true
   info: |
-    Minimum possible value for auto-tuned recovery_sleep_us. Values lower
+    Minimum possible value for auto-tuned recovery_sleep_us. Lower values
-    than this value are changed to 0.
+    are changed to 0.
   info_ru: |
     Минимальное возможное значение авто-подстроенного recovery_sleep_us.
-    Значения ниже данного заменяются на 0.
+    Меньшие значения заменяются на 0.
+- name: recovery_tune_sleep_cutoff_us
+  type: us
+  default: 10000000
+  online: true
+  info: |
+    Maximum possible value for auto-tuned recovery_sleep_us. Higher values
+    are treated as outliers and ignored in aggregation.
+  info_ru: |
+    Максимальное возможное значение авто-подстроенного recovery_sleep_us.
+    Большие значения считаются случайными выбросами и игнорируются в
+    усреднении.

docs/usage/cli.en.md

@@ -131,19 +131,18 @@ See also about [how to export snapshots](qemu.en.md#exporting-snapshots).
 
 ## modify
 
-`vitastor-cli modify <name> [--rename <new-name>] [--resize <size>] [--readonly | --readwrite] [-f|--force]`
+`vitastor-cli modify <name> [--rename <new-name>] [--resize <size>] [--readonly | --readwrite] [-f|--force] [--down-ok]`
 
 Rename, resize image or change its readonly status. Images with children can't be made read-write.
 If the new size is smaller than the old size, extra data will be purged.
 You should resize file system in the image, if present, before shrinking it.
 
-```
+| `-f|--force` | Proceed with shrinking or setting readwrite flag even if the image has children. |
--f|--force  Proceed with shrinking or setting readwrite flag even if the image has children.
+| `--down-ok`  | Proceed with shrinking even if some data will be left on unavailable OSDs.       |
-```
 
 ## rm
 
-`vitastor-cli rm <from> [<to>] [--writers-stopped]`
+`vitastor-cli rm <from> [<to>] [--writers-stopped] [--down-ok]`
 
 Remove `<from>` or all layers between `<from>` and `<to>` (`<to>` must be a child of `<from>`),
 rebasing all their children accordingly. --writers-stopped allows merging to be a bit
@@ -151,6 +150,10 @@ more effective in case of a single 'slim' read-write child and 'fat' removed par
 the child is merged into parent and parent is renamed to child in that case.
 In other cases parent layers are always merged into children.
 
+Other options:
+
+| `--down-ok` | Continue deletion/merging even if some data will be left on unavailable OSDs. |
+
 ## flatten
 
 `vitastor-cli flatten <layer>`

docs/usage/cli.ru.md

@@ -132,7 +132,7 @@ vitastor-cli snap-create [-p|--pool <id|name>] <image>@<snapshot>
 
 ## modify
 
-`vitastor-cli modify <name> [--rename <new-name>] [--resize <size>] [--readonly | --readwrite] [-f|--force]`
+`vitastor-cli modify <name> [--rename <new-name>] [--resize <size>] [--readonly | --readwrite] [-f|--force] [--down-ok]`
 
 Изменить размер, имя образа или флаг "только для чтения". Снимать флаг "только для чтения"
 и уменьшать размер образов, у которых есть дочерние клоны, без `--force` нельзя.
@@ -140,13 +140,12 @@ vitastor-cli snap-create [-p|--pool <id|name>] <image>@<snapshot>
 Если новый размер меньше старого, "лишние" данные будут удалены, поэтому перед уменьшением
 образа сначала уменьшите файловую систему в нём.
 
-```
+| -f|--force | Разрешить уменьшение или перевод в чтение-запись образа, у которого есть клоны.        |
--f|--force  Разрешить уменьшение или перевод в чтение-запись образа, у которого есть клоны.
+| --down-ok  | Разрешить уменьшение, даже если часть данных останется неудалённой на недоступных OSD. |
-```
 
 ## rm
 
-`vitastor-cli rm <from> [<to>] [--writers-stopped]`
+`vitastor-cli rm <from> [<to>] [--writers-stopped] [--down-ok]`
 
 Удалить образ `<from>` или все слои от `<from>` до `<to>` (`<to>` должен быть дочерним
 образом `<from>`), одновременно меняя родительские образы их клонов (если таковые есть).
@@ -158,6 +157,10 @@ vitastor-cli snap-create [-p|--pool <id|name>] <image>@<snapshot>
 
 В других случаях родительские слои вливаются в дочерние.
 
+Другие опции:
+
+| `--down-ok` | Продолжать удаление/слияние, даже если часть данных останется неудалённой на недоступных OSD. |
+
 ## flatten
 
 `vitastor-cli flatten <layer>`

docs/usage/disk.en.md

@@ -261,7 +261,7 @@ Options (see also [Cluster-Wide Disk Layout Parameters](../config/layout-cluster
 ```
 --object_size 128k       Set blockstore block size
 --bitmap_granularity 4k  Set bitmap granularity
---journal_size 16M       Set journal size
+--journal_size 32M       Set journal size
 --data_csum_type none    Set data checksum type (crc32c or none)
 --csum_block_size 4k     Set data checksum block size
 --device_block_size 4k   Set device block size

docs/usage/disk.ru.md

@@ -267,7 +267,7 @@ OSD отключены fsync-и.
 ```
 --object_size 128k       Размер блока хранилища
 --bitmap_granularity 4k  Гранулярность битовых карт
---journal_size 16M       Размер журнала
+--journal_size 32M       Размер журнала
 --data_csum_type none    Задать тип контрольных сумм (crc32c или none)
 --csum_block_size 4k     Задать размер блока расчёта контрольных сумм
 --device_block_size 4k   Размер блока устройства

mon/mon.js

@@ -86,13 +86,14 @@ const etcd_tree = {
             client_max_buffered_bytes: 33554432,
             client_max_buffered_ops: 1024,
             client_max_writeback_iodepth: 256,
+            client_retry_interval: 50, // ms. min: 10
+            client_eio_retry_interval: 1000, // ms
             // client and osd - configurable online
             log_level: 0,
             peer_connect_interval: 5, // seconds. min: 1
             peer_connect_timeout: 5, // seconds. min: 1
             osd_idle_timeout: 5, // seconds. min: 1
             osd_ping_timeout: 5, // seconds. min: 1
-            up_wait_retry_interval: 50, // ms. min: 10
             max_etcd_attempts: 5,
             etcd_quick_timeout: 1000, // ms
             etcd_slow_timeout: 5000, // ms
@@ -675,7 +676,12 @@ class Mon
                 {
                     this.parse_kv(e.kv);
                     const key = e.kv.key.substr(this.etcd_prefix.length);
-                    if (key.substr(0, 11) == '/osd/stats/' || key.substr(0, 10) == '/pg/stats/' || key.substr(0, 16) == '/osd/inodestats/')
+                    if (key.substr(0, 11) == '/osd/state/')
+                    {
+                        stats_changed = true;
+                        changed = true;
+                    }
+                    else if (key.substr(0, 11) == '/osd/stats/' || key.substr(0, 10) == '/pg/stats/' || key.substr(0, 16) == '/osd/inodestats/')
                     {
                         stats_changed = true;
                     }
@@ -1635,9 +1641,13 @@ class Mon
         }
         const sum_diff = { op_stats: {}, subop_stats: {}, recovery_stats: {} };
         // Sum derived values instead of deriving summed
-        for (const osd in this.state.osd.stats)
+        for (const osd in this.state.osd.state)
         {
             const derived = this.prev_stats.osd_diff[osd];
+            if (!this.state.osd.state[osd] || !derived)
+            {
+                continue;
+            }
             for (const type in sum_diff)
             {
                 for (const op in derived[type]||{})
@@ -1738,9 +1748,13 @@ class Mon
             const used = this.state.pool.stats[pool_id].used_raw_tb;
             this.state.pool.stats[pool_id].used_raw_tb = Number(used)/1024/1024/1024/1024;
         }
-        for (const osd_num in this.state.osd.inodestats)
+        for (const osd_num in this.state.osd.state)
         {
             const ist = this.state.osd.inodestats[osd_num];
+            if (!ist || !this.state.osd.state[osd_num])
+            {
+                continue;
+            }
             for (const pool_id in ist)
             {
                 inode_stats[pool_id] = inode_stats[pool_id] || {};
@@ -1756,9 +1770,14 @@ class Mon
                 }
             }
         }
-        for (const osd in this.prev_stats.osd_diff)
+        for (const osd in this.state.osd.state)
         {
-            for (const pool_id in this.prev_stats.osd_diff[osd].inode_stats)
+            const osd_diff = this.prev_stats.osd_diff[osd];
+            if (!osd_diff || !this.state.osd.state[osd])
+            {
+                continue;
+            }
+            for (const pool_id in osd_diff.inode_stats)
             {
                 for (const inode_num in this.prev_stats.osd_diff[osd].inode_stats[pool_id])
                 {

mon/package.json

@@ -1,6 +1,6 @@
 {
   "name": "vitastor-mon",
-  "version": "1.4.2",
+  "version": "1.4.8",
   "description": "Vitastor SDS monitor service",
   "main": "mon-main.js",
   "scripts": {

patches/cinder-vitastor.py

@@ -50,7 +50,7 @@ from cinder.volume import configuration
 from cinder.volume import driver
 from cinder.volume import volume_utils
 
-VERSION = '1.4.2'
+VERSION = '1.4.8'
 
 LOG = logging.getLogger(__name__)
 

patches/libvirt-8.0-vitastor.diff

@@ -0,0 +1,692 @@
+commit d85024bd803b3b91f15578ed22de4ce31856626f
+Author: Vitaliy Filippov <vitalif@yourcmc.ru>
+Date:   Wed Jan 24 18:07:43 2024 +0300
+
+    Add Vitastor support
+
+diff --git a/docs/schemas/domaincommon.rng b/docs/schemas/domaincommon.rng
+index 7fa5c2b8b5..2d77f391e7 100644
+--- a/docs/schemas/domaincommon.rng
++++ b/docs/schemas/domaincommon.rng
+@@ -1898,6 +1898,35 @@
+     </element>
+   </define>
+ 
++  <define name="diskSourceNetworkProtocolVitastor">
++    <element name="source">
++      <interleave>
++        <attribute name="protocol">
++          <value>vitastor</value>
++        </attribute>
++        <ref name="diskSourceCommon"/>
++        <optional>
++          <attribute name="name"/>
++        </optional>
++        <optional>
++          <attribute name="query"/>
++        </optional>
++        <zeroOrMore>
++          <ref name="diskSourceNetworkHost"/>
++        </zeroOrMore>
++        <optional>
++          <element name="config">
++            <attribute name="file">
++              <ref name="absFilePath"/>
++            </attribute>
++            <empty/>
++          </element>
++        </optional>
++        <empty/>
++      </interleave>
++    </element>
++  </define>
++
+   <define name="diskSourceNetworkProtocolISCSI">
+     <element name="source">
+       <attribute name="protocol">
+@@ -2154,6 +2183,7 @@
+       <ref name="diskSourceNetworkProtocolSimple"/>
+       <ref name="diskSourceNetworkProtocolVxHS"/>
+       <ref name="diskSourceNetworkProtocolNFS"/>
++      <ref name="diskSourceNetworkProtocolVitastor"/>
+     </choice>
+   </define>
+ 
+diff --git a/include/libvirt/libvirt-storage.h b/include/libvirt/libvirt-storage.h
+index f89856b93e..a8cb9387e2 100644
+--- a/include/libvirt/libvirt-storage.h
++++ b/include/libvirt/libvirt-storage.h
+@@ -246,6 +246,7 @@ typedef enum {
+     VIR_CONNECT_LIST_STORAGE_POOLS_ZFS           = 1 << 17,
+     VIR_CONNECT_LIST_STORAGE_POOLS_VSTORAGE      = 1 << 18,
+     VIR_CONNECT_LIST_STORAGE_POOLS_ISCSI_DIRECT  = 1 << 19,
++    VIR_CONNECT_LIST_STORAGE_POOLS_VITASTOR      = 1 << 20,
+ } virConnectListAllStoragePoolsFlags;
+ 
+ int                     virConnectListAllStoragePools(virConnectPtr conn,
+diff --git a/src/conf/domain_conf.c b/src/conf/domain_conf.c
+index 5691b8d2d5..6669e8451d 100644
+--- a/src/conf/domain_conf.c
++++ b/src/conf/domain_conf.c
+@@ -8293,7 +8293,8 @@ virDomainDiskSourceNetworkParse(xmlNodePtr node,
+     src->configFile = virXPathString("string(./config/@file)", ctxt);
+ 
+     if (src->protocol == VIR_STORAGE_NET_PROTOCOL_HTTP ||
+-        src->protocol == VIR_STORAGE_NET_PROTOCOL_HTTPS)
++        src->protocol == VIR_STORAGE_NET_PROTOCOL_HTTPS ||
++        src->protocol == VIR_STORAGE_NET_PROTOCOL_VITASTOR)
+         src->query = virXMLPropString(node, "query");
+ 
+     if (virDomainStorageNetworkParseHosts(node, ctxt, &src->hosts, &src->nhosts) < 0)
+@@ -31267,6 +31268,7 @@ virDomainStorageSourceTranslateSourcePool(virStorageSource *src,
+ 
+     case VIR_STORAGE_POOL_MPATH:
+     case VIR_STORAGE_POOL_RBD:
++    case VIR_STORAGE_POOL_VITASTOR:
+     case VIR_STORAGE_POOL_SHEEPDOG:
+     case VIR_STORAGE_POOL_GLUSTER:
+     case VIR_STORAGE_POOL_LAST:
+diff --git a/src/conf/domain_validate.c b/src/conf/domain_validate.c
+index a4271f1247..621c1b7b31 100644
+--- a/src/conf/domain_validate.c
++++ b/src/conf/domain_validate.c
+@@ -508,7 +508,7 @@ virDomainDiskDefValidateSourceChainOne(const virStorageSource *src)
+         }
+     }
+ 
+-    /* internal snapshots and config files are currently supported only with rbd: */
++    /* internal snapshots are currently supported only with rbd: */
+     if (virStorageSourceGetActualType(src) != VIR_STORAGE_TYPE_NETWORK &&
+         src->protocol != VIR_STORAGE_NET_PROTOCOL_RBD) {
+         if (src->snapshot) {
+@@ -517,11 +517,15 @@ virDomainDiskDefValidateSourceChainOne(const virStorageSource *src)
+                              "only with 'rbd' disks"));
+             return -1;
+         }
+-
++    }
++    /* config files are currently supported only with rbd and vitastor: */
++    if (virStorageSourceGetActualType(src) != VIR_STORAGE_TYPE_NETWORK &&
++        src->protocol != VIR_STORAGE_NET_PROTOCOL_RBD &&
++        src->protocol != VIR_STORAGE_NET_PROTOCOL_VITASTOR) {
+         if (src->configFile) {
+             virReportError(VIR_ERR_XML_ERROR, "%s",
+                            _("<config> element is currently supported "
+-                             "only with 'rbd' disks"));
++                             "only with 'rbd' and 'vitastor' disks"));
+             return -1;
+         }
+     }
+diff --git a/src/conf/storage_conf.c b/src/conf/storage_conf.c
+index 6690d26ffd..2255df9d28 100644
+--- a/src/conf/storage_conf.c
++++ b/src/conf/storage_conf.c
+@@ -60,7 +60,7 @@ VIR_ENUM_IMPL(virStoragePool,
+               "logical", "disk", "iscsi",
+               "iscsi-direct", "scsi", "mpath",
+               "rbd", "sheepdog", "gluster",
+-              "zfs", "vstorage",
++              "zfs", "vstorage", "vitastor",
+ );
+ 
+ VIR_ENUM_IMPL(virStoragePoolFormatFileSystem,
+@@ -246,6 +246,18 @@ static virStoragePoolTypeInfo poolTypeInfo[] = {
+           .formatToString = virStorageFileFormatTypeToString,
+       }
+     },
++    {.poolType = VIR_STORAGE_POOL_VITASTOR,
++     .poolOptions = {
++         .flags = (VIR_STORAGE_POOL_SOURCE_HOST |
++                   VIR_STORAGE_POOL_SOURCE_NETWORK |
++                   VIR_STORAGE_POOL_SOURCE_NAME),
++      },
++      .volOptions = {
++          .defaultFormat = VIR_STORAGE_FILE_RAW,
++          .formatFromString = virStorageVolumeFormatFromString,
++          .formatToString = virStorageFileFormatTypeToString,
++      }
++    },
+     {.poolType = VIR_STORAGE_POOL_SHEEPDOG,
+      .poolOptions = {
+          .flags = (VIR_STORAGE_POOL_SOURCE_HOST |
+@@ -546,6 +558,11 @@ virStoragePoolDefParseSource(xmlXPathContextPtr ctxt,
+                        _("element 'name' is mandatory for RBD pool"));
+         return -1;
+     }
++    if (pool_type == VIR_STORAGE_POOL_VITASTOR && source->name == NULL) {
++        virReportError(VIR_ERR_XML_ERROR, "%s",
++                       _("element 'name' is mandatory for Vitastor pool"));
++        return -1;
++    }
+ 
+     if (options->formatFromString) {
+         g_autofree char *format = NULL;
+@@ -1176,6 +1193,7 @@ virStoragePoolDefFormatBuf(virBuffer *buf,
+     /* RBD, Sheepdog, Gluster and Iscsi-direct devices are not local block devs nor
+      * files, so they don't have a target */
+     if (def->type != VIR_STORAGE_POOL_RBD &&
++        def->type != VIR_STORAGE_POOL_VITASTOR &&
+         def->type != VIR_STORAGE_POOL_SHEEPDOG &&
+         def->type != VIR_STORAGE_POOL_GLUSTER &&
+         def->type != VIR_STORAGE_POOL_ISCSI_DIRECT) {
+diff --git a/src/conf/storage_conf.h b/src/conf/storage_conf.h
+index aaecf138d6..97172db38b 100644
+--- a/src/conf/storage_conf.h
++++ b/src/conf/storage_conf.h
+@@ -106,6 +106,7 @@ typedef enum {
+     VIR_STORAGE_POOL_GLUSTER,  /* Gluster device */
+     VIR_STORAGE_POOL_ZFS,      /* ZFS */
+     VIR_STORAGE_POOL_VSTORAGE, /* Virtuozzo Storage */
++    VIR_STORAGE_POOL_VITASTOR, /* Vitastor */
+ 
+     VIR_STORAGE_POOL_LAST,
+ } virStoragePoolType;
+@@ -466,6 +467,7 @@ VIR_ENUM_DECL(virStoragePartedFs);
+                  VIR_CONNECT_LIST_STORAGE_POOLS_SCSI     | \
+                  VIR_CONNECT_LIST_STORAGE_POOLS_MPATH    | \
+                  VIR_CONNECT_LIST_STORAGE_POOLS_RBD      | \
++                 VIR_CONNECT_LIST_STORAGE_POOLS_VITASTOR | \
+                  VIR_CONNECT_LIST_STORAGE_POOLS_SHEEPDOG | \
+                  VIR_CONNECT_LIST_STORAGE_POOLS_GLUSTER  | \
+                  VIR_CONNECT_LIST_STORAGE_POOLS_ZFS      | \
+diff --git a/src/conf/storage_source_conf.c b/src/conf/storage_source_conf.c
+index d42f715f26..29d8da3d10 100644
+--- a/src/conf/storage_source_conf.c
++++ b/src/conf/storage_source_conf.c
+@@ -86,6 +86,7 @@ VIR_ENUM_IMPL(virStorageNetProtocol,
+               "ssh",
+               "vxhs",
+               "nfs",
++              "vitastor",
+ );
+ 
+ 
+@@ -1265,6 +1266,7 @@ virStorageSourceNetworkDefaultPort(virStorageNetProtocol protocol)
+         case VIR_STORAGE_NET_PROTOCOL_GLUSTER:
+             return 24007;
+ 
++        case VIR_STORAGE_NET_PROTOCOL_VITASTOR:
+         case VIR_STORAGE_NET_PROTOCOL_RBD:
+             /* we don't provide a default for RBD */
+             return 0;
+diff --git a/src/conf/storage_source_conf.h b/src/conf/storage_source_conf.h
+index c4a026881c..67568e9181 100644
+--- a/src/conf/storage_source_conf.h
++++ b/src/conf/storage_source_conf.h
+@@ -128,6 +128,7 @@ typedef enum {
+     VIR_STORAGE_NET_PROTOCOL_SSH,
+     VIR_STORAGE_NET_PROTOCOL_VXHS,
+     VIR_STORAGE_NET_PROTOCOL_NFS,
++    VIR_STORAGE_NET_PROTOCOL_VITASTOR,
+ 
+     VIR_STORAGE_NET_PROTOCOL_LAST
+ } virStorageNetProtocol;
+diff --git a/src/conf/virstorageobj.c b/src/conf/virstorageobj.c
+index 02903ac487..504df599fb 100644
+--- a/src/conf/virstorageobj.c
++++ b/src/conf/virstorageobj.c
+@@ -1481,6 +1481,7 @@ virStoragePoolObjSourceFindDuplicateCb(const void *payload,
+             return 1;
+         break;
+ 
++    case VIR_STORAGE_POOL_VITASTOR:
+     case VIR_STORAGE_POOL_RBD:
+     case VIR_STORAGE_POOL_LAST:
+         break;
+@@ -1978,6 +1979,8 @@ virStoragePoolObjMatch(virStoragePoolObj *obj,
+                (obj->def->type == VIR_STORAGE_POOL_MPATH))   ||
+               (MATCH(VIR_CONNECT_LIST_STORAGE_POOLS_RBD) &&
+                (obj->def->type == VIR_STORAGE_POOL_RBD))     ||
++              (MATCH(VIR_CONNECT_LIST_STORAGE_POOLS_VITASTOR) &&
++               (obj->def->type == VIR_STORAGE_POOL_VITASTOR)) ||
+               (MATCH(VIR_CONNECT_LIST_STORAGE_POOLS_SHEEPDOG) &&
+                (obj->def->type == VIR_STORAGE_POOL_SHEEPDOG)) ||
+               (MATCH(VIR_CONNECT_LIST_STORAGE_POOLS_GLUSTER) &&
+diff --git a/src/libvirt-storage.c b/src/libvirt-storage.c
+index cbc522b300..b4760fa58d 100644
+--- a/src/libvirt-storage.c
++++ b/src/libvirt-storage.c
+@@ -92,6 +92,7 @@ virStoragePoolGetConnect(virStoragePoolPtr pool)
+  * VIR_CONNECT_LIST_STORAGE_POOLS_SCSI
+  * VIR_CONNECT_LIST_STORAGE_POOLS_MPATH
+  * VIR_CONNECT_LIST_STORAGE_POOLS_RBD
++ * VIR_CONNECT_LIST_STORAGE_POOLS_VITASTOR
+  * VIR_CONNECT_LIST_STORAGE_POOLS_SHEEPDOG
+  * VIR_CONNECT_LIST_STORAGE_POOLS_GLUSTER
+  * VIR_CONNECT_LIST_STORAGE_POOLS_ZFS
+diff --git a/src/libxl/libxl_conf.c b/src/libxl/libxl_conf.c
+index 1ac6253ad7..abe4587f94 100644
+--- a/src/libxl/libxl_conf.c
++++ b/src/libxl/libxl_conf.c
+@@ -962,6 +962,7 @@ libxlMakeNetworkDiskSrcStr(virStorageSource *src,
+     case VIR_STORAGE_NET_PROTOCOL_SSH:
+     case VIR_STORAGE_NET_PROTOCOL_VXHS:
+     case VIR_STORAGE_NET_PROTOCOL_NFS:
++    case VIR_STORAGE_NET_PROTOCOL_VITASTOR:
+     case VIR_STORAGE_NET_PROTOCOL_LAST:
+     case VIR_STORAGE_NET_PROTOCOL_NONE:
+         virReportError(VIR_ERR_NO_SUPPORT,
+diff --git a/src/libxl/xen_xl.c b/src/libxl/xen_xl.c
+index 7604e3d534..6453bb9776 100644
+--- a/src/libxl/xen_xl.c
++++ b/src/libxl/xen_xl.c
+@@ -1506,6 +1506,7 @@ xenFormatXLDiskSrcNet(virStorageSource *src)
+     case VIR_STORAGE_NET_PROTOCOL_SSH:
+     case VIR_STORAGE_NET_PROTOCOL_VXHS:
+     case VIR_STORAGE_NET_PROTOCOL_NFS:
++    case VIR_STORAGE_NET_PROTOCOL_VITASTOR:
+     case VIR_STORAGE_NET_PROTOCOL_LAST:
+     case VIR_STORAGE_NET_PROTOCOL_NONE:
+         virReportError(VIR_ERR_NO_SUPPORT,
+diff --git a/src/qemu/qemu_block.c b/src/qemu/qemu_block.c
+index e5ff653a60..884ecc79ea 100644
+--- a/src/qemu/qemu_block.c
++++ b/src/qemu/qemu_block.c
+@@ -943,6 +943,38 @@ qemuBlockStorageSourceGetRBDProps(virStorageSource *src,
+ }
+ 
+ 
++static virJSONValue *
++qemuBlockStorageSourceGetVitastorProps(virStorageSource *src)
++{
++    virJSONValue *ret = NULL;
++    virStorageNetHostDef *host;
++    size_t i;
++    g_auto(virBuffer) buf = VIR_BUFFER_INITIALIZER;
++    g_autofree char *etcd = NULL;
++
++    for (i = 0; i < src->nhosts; i++) {
++        host = src->hosts + i;
++        if ((virStorageNetHostTransport)host->transport != VIR_STORAGE_NET_HOST_TRANS_TCP) {
++            return NULL;
++        }
++        virBufferAsprintf(&buf, i > 0 ? ",%s:%u" : "%s:%u", host->name, host->port);
++    }
++    if (src->nhosts > 0) {
++        etcd = virBufferContentAndReset(&buf);
++    }
++
++    if (virJSONValueObjectCreate(&ret,
++                                 "S:etcd-host", etcd,
++                                 "S:etcd-prefix", src->query,
++                                 "S:config-path", src->configFile,
++                                 "s:image", src->path,
++                                 NULL) < 0)
++        return NULL;
++
++    return ret;
++}
++
++
+ static virJSONValue *
+ qemuBlockStorageSourceGetSheepdogProps(virStorageSource *src)
+ {
+@@ -1233,6 +1265,12 @@ qemuBlockStorageSourceGetBackendProps(virStorageSource *src,
+                 return NULL;
+             break;
+ 
++        case VIR_STORAGE_NET_PROTOCOL_VITASTOR:
++            driver = "vitastor";
++            if (!(fileprops = qemuBlockStorageSourceGetVitastorProps(src)))
++                return NULL;
++            break;
++
+         case VIR_STORAGE_NET_PROTOCOL_SHEEPDOG:
+             driver = "sheepdog";
+             if (!(fileprops = qemuBlockStorageSourceGetSheepdogProps(src)))
+@@ -2244,6 +2282,7 @@ qemuBlockGetBackingStoreString(virStorageSource *src,
+ 
+             case VIR_STORAGE_NET_PROTOCOL_SHEEPDOG:
+             case VIR_STORAGE_NET_PROTOCOL_RBD:
++            case VIR_STORAGE_NET_PROTOCOL_VITASTOR:
+             case VIR_STORAGE_NET_PROTOCOL_VXHS:
+             case VIR_STORAGE_NET_PROTOCOL_NFS:
+             case VIR_STORAGE_NET_PROTOCOL_SSH:
+@@ -2626,6 +2665,12 @@ qemuBlockStorageSourceCreateGetStorageProps(virStorageSource *src,
+                 return -1;
+             break;
+ 
++        case VIR_STORAGE_NET_PROTOCOL_VITASTOR:
++            driver = "vitastor";
++            if (!(location = qemuBlockStorageSourceGetVitastorProps(src)))
++                return -1;
++            break;
++
+         case VIR_STORAGE_NET_PROTOCOL_SHEEPDOG:
+             driver = "sheepdog";
+             if (!(location = qemuBlockStorageSourceGetSheepdogProps(src)))
+diff --git a/src/qemu/qemu_command.c b/src/qemu/qemu_command.c
+index d822533ccb..afe2087303 100644
+--- a/src/qemu/qemu_command.c
++++ b/src/qemu/qemu_command.c
+@@ -1723,6 +1723,43 @@ qemuBuildNetworkDriveStr(virStorageSource *src,
+             ret = virBufferContentAndReset(&buf);
+             break;
+ 
++        case VIR_STORAGE_NET_PROTOCOL_VITASTOR:
++            if (strchr(src->path, ':')) {
++                virReportError(VIR_ERR_CONFIG_UNSUPPORTED,
++                               _("':' not allowed in Vitastor source volume name '%s'"),
++                               src->path);
++                return NULL;
++            }
++
++            virBufferStrcat(&buf, "vitastor:image=", src->path, NULL);
++
++            if (src->nhosts > 0) {
++                virBufferAddLit(&buf, ":etcd-host=");
++                for (i = 0; i < src->nhosts; i++) {
++                    if (i)
++                        virBufferAddLit(&buf, ",");
++
++                    /* assume host containing : is ipv6 */
++                    if (strchr(src->hosts[i].name, ':'))
++                        virBufferEscape(&buf, '\\', ":", "[%s]",
++                                        src->hosts[i].name);
++                    else
++                        virBufferAsprintf(&buf, "%s", src->hosts[i].name);
++
++                    if (src->hosts[i].port)
++                        virBufferAsprintf(&buf, "\\:%u", src->hosts[i].port);
++                }
++            }
++
++            if (src->configFile)
++                virBufferEscape(&buf, '\\', ":", ":config-path=%s", src->configFile);
++
++            if (src->query)
++                virBufferEscape(&buf, '\\', ":", ":etcd-prefix=%s", src->query);
++
++            ret = virBufferContentAndReset(&buf);
++            break;
++
+         case VIR_STORAGE_NET_PROTOCOL_VXHS:
+             virReportError(VIR_ERR_INTERNAL_ERROR, "%s",
+                            _("VxHS protocol does not support URI syntax"));
+diff --git a/src/qemu/qemu_domain.c b/src/qemu/qemu_domain.c
+index a8401bac30..3dc1fe6db0 100644
+--- a/src/qemu/qemu_domain.c
++++ b/src/qemu/qemu_domain.c
+@@ -4731,7 +4731,8 @@ qemuDomainValidateStorageSource(virStorageSource *src,
+     if (src->query &&
+         (actualType != VIR_STORAGE_TYPE_NETWORK ||
+          (src->protocol != VIR_STORAGE_NET_PROTOCOL_HTTPS &&
+-          src->protocol != VIR_STORAGE_NET_PROTOCOL_HTTP))) {
++          src->protocol != VIR_STORAGE_NET_PROTOCOL_HTTP &&
++          src->protocol != VIR_STORAGE_NET_PROTOCOL_VITASTOR))) {
+         virReportError(VIR_ERR_CONFIG_UNSUPPORTED, "%s",
+                        _("query is supported only with HTTP(S) protocols"));
+         return -1;
+@@ -9919,6 +9920,7 @@ qemuDomainPrepareStorageSourceTLS(virStorageSource *src,
+         break;
+ 
+     case VIR_STORAGE_NET_PROTOCOL_RBD:
++    case VIR_STORAGE_NET_PROTOCOL_VITASTOR:
+     case VIR_STORAGE_NET_PROTOCOL_SHEEPDOG:
+     case VIR_STORAGE_NET_PROTOCOL_GLUSTER:
+     case VIR_STORAGE_NET_PROTOCOL_ISCSI:
+diff --git a/src/qemu/qemu_snapshot.c b/src/qemu/qemu_snapshot.c
+index f92e00f9c0..854a3fbc90 100644
+--- a/src/qemu/qemu_snapshot.c
++++ b/src/qemu/qemu_snapshot.c
+@@ -393,6 +393,7 @@ qemuSnapshotPrepareDiskExternalInactive(virDomainSnapshotDiskDef *snapdisk,
+         case VIR_STORAGE_NET_PROTOCOL_NONE:
+         case VIR_STORAGE_NET_PROTOCOL_NBD:
+         case VIR_STORAGE_NET_PROTOCOL_RBD:
++        case VIR_STORAGE_NET_PROTOCOL_VITASTOR:
+         case VIR_STORAGE_NET_PROTOCOL_SHEEPDOG:
+         case VIR_STORAGE_NET_PROTOCOL_GLUSTER:
+         case VIR_STORAGE_NET_PROTOCOL_ISCSI:
+@@ -485,6 +486,7 @@ qemuSnapshotPrepareDiskExternalActive(virDomainObj *vm,
+         case VIR_STORAGE_NET_PROTOCOL_NONE:
+         case VIR_STORAGE_NET_PROTOCOL_NBD:
+         case VIR_STORAGE_NET_PROTOCOL_RBD:
++        case VIR_STORAGE_NET_PROTOCOL_VITASTOR:
+         case VIR_STORAGE_NET_PROTOCOL_SHEEPDOG:
+         case VIR_STORAGE_NET_PROTOCOL_ISCSI:
+         case VIR_STORAGE_NET_PROTOCOL_HTTP:
+@@ -638,6 +640,7 @@ qemuSnapshotPrepareDiskInternal(virDomainDiskDef *disk,
+         case VIR_STORAGE_NET_PROTOCOL_NONE:
+         case VIR_STORAGE_NET_PROTOCOL_NBD:
+         case VIR_STORAGE_NET_PROTOCOL_RBD:
++        case VIR_STORAGE_NET_PROTOCOL_VITASTOR:
+         case VIR_STORAGE_NET_PROTOCOL_SHEEPDOG:
+         case VIR_STORAGE_NET_PROTOCOL_GLUSTER:
+         case VIR_STORAGE_NET_PROTOCOL_ISCSI:
+diff --git a/src/storage/storage_driver.c b/src/storage/storage_driver.c
+index 4df2c75a2b..5a5e48ef71 100644
+--- a/src/storage/storage_driver.c
++++ b/src/storage/storage_driver.c
+@@ -1643,6 +1643,7 @@ storageVolLookupByPathCallback(virStoragePoolObj *obj,
+ 
+         case VIR_STORAGE_POOL_GLUSTER:
+         case VIR_STORAGE_POOL_RBD:
++        case VIR_STORAGE_POOL_VITASTOR:
+         case VIR_STORAGE_POOL_SHEEPDOG:
+         case VIR_STORAGE_POOL_ZFS:
+         case VIR_STORAGE_POOL_LAST:
+diff --git a/src/storage_file/storage_source_backingstore.c b/src/storage_file/storage_source_backingstore.c
+index e48ae725ab..2017ccc88c 100644
+--- a/src/storage_file/storage_source_backingstore.c
++++ b/src/storage_file/storage_source_backingstore.c
+@@ -284,6 +284,75 @@ virStorageSourceParseRBDColonString(const char *rbdstr,
+ }
+ 
+ 
++static int
++virStorageSourceParseVitastorColonString(const char *colonstr,
++                                         virStorageSource *src)
++{
++    char *p, *e, *next;
++    g_autofree char *options = NULL;
++
++    /* optionally skip the "vitastor:" prefix if provided */
++    if (STRPREFIX(colonstr, "vitastor:"))
++        colonstr += strlen("vitastor:");
++
++    options = g_strdup(colonstr);
++
++    p = options;
++    while (*p) {
++        /* find : delimiter or end of string */
++        for (e = p; *e && *e != ':'; ++e) {
++            if (*e == '\\') {
++                e++;
++                if (*e == '\0')
++                    break;
++            }
++        }
++        if (*e == '\0') {
++            next = e;    /* last kv pair */
++        } else {
++            next = e + 1;
++            *e = '\0';
++        }
++
++        if (STRPREFIX(p, "image=")) {
++            src->path = g_strdup(p + strlen("image="));
++        } else if (STRPREFIX(p, "etcd-prefix=")) {
++            src->query = g_strdup(p + strlen("etcd-prefix="));
++        } else if (STRPREFIX(p, "config-path=")) {
++            src->configFile = g_strdup(p + strlen("config-path="));
++        } else if (STRPREFIX(p, "etcd-host=")) {
++            char *h, *sep;
++
++            h = p + strlen("etcd-host=");
++            while (h < e) {
++                for (sep = h; sep < e; ++sep) {
++                    if (*sep == '\\' && (sep[1] == ',' ||
++                                         sep[1] == ';' ||
++                                         sep[1] == ' ')) {
++                        *sep = '\0';
++                        sep += 2;
++                        break;
++                    }
++                }
++
++                if (virStorageSourceRBDAddHost(src, h) < 0)
++                    return -1;
++
++                h = sep;
++            }
++        }
++
++        p = next;
++    }
++
++    if (!src->path) {
++        return -1;
++    }
++
++    return 0;
++}
++
++
+ static int
+ virStorageSourceParseNBDColonString(const char *nbdstr,
+                                     virStorageSource *src)
+@@ -396,6 +465,11 @@ virStorageSourceParseBackingColon(virStorageSource *src,
+             return -1;
+         break;
+ 
++    case VIR_STORAGE_NET_PROTOCOL_VITASTOR:
++        if (virStorageSourceParseVitastorColonString(path, src) < 0)
++            return -1;
++        break;
++
+     case VIR_STORAGE_NET_PROTOCOL_SHEEPDOG:
+     case VIR_STORAGE_NET_PROTOCOL_LAST:
+     case VIR_STORAGE_NET_PROTOCOL_NONE:
+@@ -984,6 +1058,54 @@ virStorageSourceParseBackingJSONRBD(virStorageSource *src,
+     return 0;
+ }
+ 
++static int
++virStorageSourceParseBackingJSONVitastor(virStorageSource *src,
++                                         virJSONValue *json,
++                                         const char *jsonstr G_GNUC_UNUSED,
++                                         int opaque G_GNUC_UNUSED)
++{
++    const char *filename;
++    const char *image = virJSONValueObjectGetString(json, "image");
++    const char *conf = virJSONValueObjectGetString(json, "config-path");
++    const char *etcd_prefix = virJSONValueObjectGetString(json, "etcd-prefix");
++    virJSONValue *servers = virJSONValueObjectGetArray(json, "server");
++    size_t nservers;
++    size_t i;
++
++    src->type = VIR_STORAGE_TYPE_NETWORK;
++    src->protocol = VIR_STORAGE_NET_PROTOCOL_VITASTOR;
++
++    /* legacy syntax passed via 'filename' option */
++    if ((filename = virJSONValueObjectGetString(json, "filename")))
++        return virStorageSourceParseVitastorColonString(filename, src);
++
++    if (!image) {
++        virReportError(VIR_ERR_INVALID_ARG, "%s",
++                       _("missing image name in Vitastor backing volume "
++                         "JSON specification"));
++        return -1;
++    }
++
++    src->path = g_strdup(image);
++    src->configFile = g_strdup(conf);
++    src->query = g_strdup(etcd_prefix);
++
++    if (servers) {
++        nservers = virJSONValueArraySize(servers);
++
++        src->hosts = g_new0(virStorageNetHostDef, nservers);
++        src->nhosts = nservers;
++
++        for (i = 0; i < nservers; i++) {
++            if (virStorageSourceParseBackingJSONInetSocketAddress(src->hosts + i,
++                                                                  virJSONValueArrayGet(servers, i)) < 0)
++                return -1;
++        }
++    }
++
++    return 0;
++}
++
+ static int
+ virStorageSourceParseBackingJSONRaw(virStorageSource *src,
+                                     virJSONValue *json,
+@@ -1162,6 +1284,7 @@ static const struct virStorageSourceJSONDriverParser jsonParsers[] = {
+     {"sheepdog", false, virStorageSourceParseBackingJSONSheepdog, 0},
+     {"ssh", false, virStorageSourceParseBackingJSONSSH, 0},
+     {"rbd", false, virStorageSourceParseBackingJSONRBD, 0},
++    {"vitastor", false, virStorageSourceParseBackingJSONVitastor, 0},
+     {"raw", true, virStorageSourceParseBackingJSONRaw, 0},
+     {"nfs", false, virStorageSourceParseBackingJSONNFS, 0},
+     {"vxhs", false, virStorageSourceParseBackingJSONVxHS, 0},
+diff --git a/src/test/test_driver.c b/src/test/test_driver.c
+index 0e93b79922..b4d33f5f56 100644
+--- a/src/test/test_driver.c
++++ b/src/test/test_driver.c
+@@ -7367,6 +7367,7 @@ testStorageVolumeTypeForPool(int pooltype)
+     case VIR_STORAGE_POOL_ISCSI_DIRECT:
+     case VIR_STORAGE_POOL_GLUSTER:
+     case VIR_STORAGE_POOL_RBD:
++    case VIR_STORAGE_POOL_VITASTOR:
+         return VIR_STORAGE_VOL_NETWORK;
+     case VIR_STORAGE_POOL_LOGICAL:
+     case VIR_STORAGE_POOL_DISK:
+diff --git a/tests/storagepoolcapsschemadata/poolcaps-fs.xml b/tests/storagepoolcapsschemadata/poolcaps-fs.xml
+index eee75af746..8bd0a57bdd 100644
+--- a/tests/storagepoolcapsschemadata/poolcaps-fs.xml
++++ b/tests/storagepoolcapsschemadata/poolcaps-fs.xml
+@@ -204,4 +204,11 @@
+       </enum>
+     </volOptions>
+   </pool>
++  <pool type='vitastor' supported='no'>
++    <volOptions>
++      <defaultFormat type='raw'/>
++      <enum name='targetFormatType'>
++      </enum>
++    </volOptions>
++  </pool>
+ </storagepoolCapabilities>
+diff --git a/tests/storagepoolcapsschemadata/poolcaps-full.xml b/tests/storagepoolcapsschemadata/poolcaps-full.xml
+index 805950a937..852df0de16 100644
+--- a/tests/storagepoolcapsschemadata/poolcaps-full.xml
++++ b/tests/storagepoolcapsschemadata/poolcaps-full.xml
+@@ -204,4 +204,11 @@
+       </enum>
+     </volOptions>
+   </pool>
++  <pool type='vitastor' supported='yes'>
++    <volOptions>
++      <defaultFormat type='raw'/>
++      <enum name='targetFormatType'>
++      </enum>
++    </volOptions>
++  </pool>
+ </storagepoolCapabilities>
+diff --git a/tests/storagepoolxml2argvtest.c b/tests/storagepoolxml2argvtest.c
+index 449b745519..7f95cc8e08 100644
+--- a/tests/storagepoolxml2argvtest.c
++++ b/tests/storagepoolxml2argvtest.c
+@@ -68,6 +68,7 @@ testCompareXMLToArgvFiles(bool shouldFail,
+     case VIR_STORAGE_POOL_GLUSTER:
+     case VIR_STORAGE_POOL_ZFS:
+     case VIR_STORAGE_POOL_VSTORAGE:
++    case VIR_STORAGE_POOL_VITASTOR:
+     case VIR_STORAGE_POOL_LAST:
+     default:
+         VIR_TEST_DEBUG("pool type '%s' has no xml2argv test", defTypeStr);
+diff --git a/tools/virsh-pool.c b/tools/virsh-pool.c
+index d391257f6e..46799c4a90 100644
+--- a/tools/virsh-pool.c
++++ b/tools/virsh-pool.c
+@@ -1213,6 +1213,9 @@ cmdPoolList(vshControl *ctl, const vshCmd *cmd G_GNUC_UNUSED)
+             case VIR_STORAGE_POOL_VSTORAGE:
+                 flags |= VIR_CONNECT_LIST_STORAGE_POOLS_VSTORAGE;
+                 break;
++            case VIR_STORAGE_POOL_VITASTOR:
++                flags |= VIR_CONNECT_LIST_STORAGE_POOLS_VITASTOR;
++                break;
+             case VIR_STORAGE_POOL_LAST:
+                 break;
+             }

rpm/build-tarball.sh

@@ -24,4 +24,4 @@ rm fio
 mv fio-copy fio
 FIO=`rpm -qi fio | perl -e 'while(<>) { /^Epoch[\s:]+(\S+)/ && print "$1:"; /^Version[\s:]+(\S+)/ && print $1; /^Release[\s:]+(\S+)/ && print "-$1"; }'`
 perl -i -pe 's/(Requires:\s*fio)([^\n]+)?/$1 = '$FIO'/' $VITASTOR/rpm/vitastor-el$EL.spec
-tar --transform 's#^#vitastor-1.4.2/#' --exclude 'rpm/*.rpm' -czf $VITASTOR/../vitastor-1.4.2$(rpm --eval '%dist').tar.gz *
+tar --transform 's#^#vitastor-1.4.8/#' --exclude 'rpm/*.rpm' -czf $VITASTOR/../vitastor-1.4.8$(rpm --eval '%dist').tar.gz *

rpm/vitastor-el7.Dockerfile

@@ -36,7 +36,7 @@ ADD . /root/vitastor
 RUN set -e; \
     cd /root/vitastor/rpm; \
     sh build-tarball.sh; \
-    cp /root/vitastor-1.4.2.el7.tar.gz ~/rpmbuild/SOURCES; \
+    cp /root/vitastor-1.4.8.el7.tar.gz ~/rpmbuild/SOURCES; \
     cp vitastor-el7.spec ~/rpmbuild/SPECS/vitastor.spec; \
     cd ~/rpmbuild/SPECS/; \
     rpmbuild -ba vitastor.spec; \

rpm/vitastor-el7.spec

@@ -1,11 +1,11 @@
 Name:           vitastor
-Version:        1.4.2
+Version:        1.4.8
 Release:        1%{?dist}
 Summary:        Vitastor, a fast software-defined clustered block storage
 
 License:        Vitastor Network Public License 1.1
 URL:            https://vitastor.io/
-Source0:        vitastor-1.4.2.el7.tar.gz
+Source0:        vitastor-1.4.8.el7.tar.gz
 
 BuildRequires:  liburing-devel >= 0.6
 BuildRequires:  gperftools-devel

rpm/vitastor-el8.Dockerfile

@@ -35,7 +35,7 @@ ADD . /root/vitastor
 RUN set -e; \
     cd /root/vitastor/rpm; \
     sh build-tarball.sh; \
-    cp /root/vitastor-1.4.2.el8.tar.gz ~/rpmbuild/SOURCES; \
+    cp /root/vitastor-1.4.8.el8.tar.gz ~/rpmbuild/SOURCES; \
     cp vitastor-el8.spec ~/rpmbuild/SPECS/vitastor.spec; \
     cd ~/rpmbuild/SPECS/; \
     rpmbuild -ba vitastor.spec; \

rpm/vitastor-el8.spec

@@ -1,11 +1,11 @@
 Name:           vitastor
-Version:        1.4.2
+Version:        1.4.8
 Release:        1%{?dist}
 Summary:        Vitastor, a fast software-defined clustered block storage
 
 License:        Vitastor Network Public License 1.1
 URL:            https://vitastor.io/
-Source0:        vitastor-1.4.2.el8.tar.gz
+Source0:        vitastor-1.4.8.el8.tar.gz
 
 BuildRequires:  liburing-devel >= 0.6
 BuildRequires:  gperftools-devel

rpm/vitastor-el9.Dockerfile

@@ -18,7 +18,7 @@ ADD . /root/vitastor
 RUN set -e; \
     cd /root/vitastor/rpm; \
     sh build-tarball.sh; \
-    cp /root/vitastor-1.4.2.el9.tar.gz ~/rpmbuild/SOURCES; \
+    cp /root/vitastor-1.4.8.el9.tar.gz ~/rpmbuild/SOURCES; \
     cp vitastor-el9.spec ~/rpmbuild/SPECS/vitastor.spec; \
     cd ~/rpmbuild/SPECS/; \
     rpmbuild -ba vitastor.spec; \

rpm/vitastor-el9.spec

@@ -1,11 +1,11 @@
 Name:           vitastor
-Version:        1.4.2
+Version:        1.4.8
 Release:        1%{?dist}
 Summary:        Vitastor, a fast software-defined clustered block storage
 
 License:        Vitastor Network Public License 1.1
 URL:            https://vitastor.io/
-Source0:        vitastor-1.4.2.el9.tar.gz
+Source0:        vitastor-1.4.8.el9.tar.gz
 
 BuildRequires:  liburing-devel >= 0.6
 BuildRequires:  gperftools-devel

src/CMakeLists.txt

@@ -16,8 +16,8 @@ if("${CMAKE_INSTALL_PREFIX}" MATCHES "^/usr/local/?$")
 	set(CMAKE_INSTALL_RPATH "${CMAKE_INSTALL_PREFIX}/${CMAKE_INSTALL_LIBDIR}")
 endif()
 
-add_definitions(-DVERSION="1.4.2")
+add_definitions(-DVERSION="1.4.8")
-add_definitions(-Wall -Wno-sign-compare -Wno-comment -Wno-parentheses -Wno-pointer-arith -fdiagnostics-color=always -fno-omit-frame-pointer -I ${CMAKE_SOURCE_DIR}/src)
+add_definitions(-D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -Wall -Wno-sign-compare -Wno-comment -Wno-parentheses -Wno-pointer-arith -fdiagnostics-color=always -fno-omit-frame-pointer -I ${CMAKE_SOURCE_DIR}/src)
 add_link_options(-fno-omit-frame-pointer)
 if (${WITH_ASAN})
 	add_definitions(-fsanitize=address)

src/blockstore_disk.cpp

@@ -108,6 +108,10 @@ void blockstore_disk_t::parse_config(std::map<std::string, std::string> & config
     {
         throw std::runtime_error("journal_block_size must be a multiple of "+std::to_string(DIRECT_IO_ALIGNMENT));
     }
+    else if (journal_block_size > MAX_DATA_BLOCK_SIZE)
+    {
+        throw std::runtime_error("journal_block_size must not exceed "+std::to_string(MAX_DATA_BLOCK_SIZE));
+    }
     if (!meta_block_size)
     {
         meta_block_size = 4096;
@@ -116,6 +120,10 @@ void blockstore_disk_t::parse_config(std::map<std::string, std::string> & config
     {
         throw std::runtime_error("meta_block_size must be a multiple of "+std::to_string(DIRECT_IO_ALIGNMENT));
     }
+    else if (meta_block_size > MAX_DATA_BLOCK_SIZE)
+    {
+        throw std::runtime_error("meta_block_size must not exceed "+std::to_string(MAX_DATA_BLOCK_SIZE));
+    }
     if (data_offset % disk_alignment)
     {
         throw std::runtime_error("data_offset must be a multiple of disk_alignment = "+std::to_string(disk_alignment));

src/blockstore_flush.cpp

@@ -19,7 +19,6 @@ journal_flusher_t::journal_flusher_t(blockstore_impl_t *bs)
     syncing_flushers = 0;
     // FIXME: allow to configure flusher_start_threshold and journal_trim_interval
     flusher_start_threshold = bs->dsk.journal_block_size / sizeof(journal_entry_stable);
-    journal_trim_interval = 512;
     journal_trim_counter = bs->journal.flush_journal ? 1 : 0;
     trim_wanted = bs->journal.flush_journal ? 1 : 0;
     journal_superblock = bs->journal.inmemory ? bs->journal.buffer : memalign_or_die(MEM_ALIGNMENT, bs->dsk.journal_block_size);
@@ -94,7 +93,7 @@ void journal_flusher_t::loop()
 void journal_flusher_t::enqueue_flush(obj_ver_id ov)
 {
 #ifdef BLOCKSTORE_DEBUG
-    printf("enqueue_flush %lx:%lx v%lu\n", ov.oid.inode, ov.oid.stripe, ov.version);
+    printf("enqueue_flush %jx:%jx v%ju\n", ov.oid.inode, ov.oid.stripe, ov.version);
 #endif
     auto it = flush_versions.find(ov.oid);
     if (it != flush_versions.end())
@@ -117,7 +116,7 @@ void journal_flusher_t::enqueue_flush(obj_ver_id ov)
 void journal_flusher_t::unshift_flush(obj_ver_id ov, bool force)
 {
 #ifdef BLOCKSTORE_DEBUG
-    printf("unshift_flush %lx:%lx v%lu\n", ov.oid.inode, ov.oid.stripe, ov.version);
+    printf("unshift_flush %jx:%jx v%ju\n", ov.oid.inode, ov.oid.stripe, ov.version);
 #endif
     auto it = flush_versions.find(ov.oid);
     if (it != flush_versions.end())
@@ -143,7 +142,7 @@ void journal_flusher_t::unshift_flush(obj_ver_id ov, bool force)
 void journal_flusher_t::remove_flush(object_id oid)
 {
 #ifdef BLOCKSTORE_DEBUG
-    printf("undo_flush %lx:%lx\n", oid.inode, oid.stripe);
+    printf("undo_flush %jx:%jx\n", oid.inode, oid.stripe);
 #endif
     auto v_it = flush_versions.find(oid);
     if (v_it != flush_versions.end())
@@ -184,8 +183,7 @@ void journal_flusher_t::mark_trim_possible()
     if (trim_wanted > 0)
     {
         dequeuing = true;
-        if (!journal_trim_counter)
+        journal_trim_counter = 0;
-            journal_trim_counter = journal_trim_interval;
         bs->ringloop->wakeup();
     }
 }
@@ -235,7 +233,7 @@ void journal_flusher_t::dump_diagnostics()
         break;
     }
     printf(
-        "Flusher: queued=%ld first=%s%lx:%lx trim_wanted=%d dequeuing=%d trimming=%d cur=%d target=%d active=%d syncing=%d\n",
+        "Flusher: queued=%zd first=%s%jx:%jx trim_wanted=%d dequeuing=%d trimming=%d cur=%d target=%d active=%d syncing=%d\n",
         flush_queue.size(), unflushable_type, unflushable.oid.inode, unflushable.oid.stripe,
         trim_wanted, dequeuing, trimming, cur_flusher_count, target_flusher_count,
         active_flushers, syncing_flushers
@@ -268,7 +266,7 @@ bool journal_flusher_t::try_find_other(std::map<obj_ver_id, dirty_entry>::iterat
 {
     int search_left = flush_queue.size() - 1;
 #ifdef BLOCKSTORE_DEBUG
-    printf("Flusher overran writers (%lx:%lx v%lu, dirty_start=%08lx) - searching for older flushes (%d left)\n",
+    printf("Flusher overran writers (%jx:%jx v%ju, dirty_start=%08jx) - searching for older flushes (%d left)\n",
         cur.oid.inode, cur.oid.stripe, cur.version, bs->journal.dirty_start, search_left);
 #endif
     while (search_left > 0)
@@ -285,7 +283,7 @@ bool journal_flusher_t::try_find_other(std::map<obj_ver_id, dirty_entry>::iterat
                 dirty_end->second.journal_sector < bs->journal.used_start))
             {
 #ifdef BLOCKSTORE_DEBUG
-                printf("Write %lx:%lx v%lu is too new: offset=%08lx\n", cur.oid.inode, cur.oid.stripe, cur.version, dirty_end->second.journal_sector);
+                printf("Write %jx:%jx v%ju is too new: offset=%08jx\n", cur.oid.inode, cur.oid.stripe, cur.version, dirty_end->second.journal_sector);
 #endif
                 enqueue_flush(cur);
             }
@@ -366,9 +364,10 @@ resume_0:
         !flusher->flush_queue.size() || !flusher->dequeuing)
     {
 stop_flusher:
-        if (flusher->trim_wanted > 0 && flusher->journal_trim_counter > 0)
+        if (flusher->trim_wanted > 0 && cur.oid.inode != 0)
         {
             // Attempt forced trim
+            cur.oid = {};
             flusher->active_flushers++;
             goto trim_journal;
         }
@@ -387,7 +386,7 @@ stop_flusher:
         if (repeat_it != flusher->sync_to_repeat.end())
         {
 #ifdef BLOCKSTORE_DEBUG
-            printf("Postpone %lx:%lx v%lu\n", cur.oid.inode, cur.oid.stripe, cur.version);
+            printf("Postpone %jx:%jx v%ju\n", cur.oid.inode, cur.oid.stripe, cur.version);
 #endif
             // We don't flush different parts of history of the same object in parallel
             // So we check if someone is already flushing this object
@@ -416,12 +415,13 @@ stop_flusher:
                 flusher->sync_to_repeat.erase(cur.oid);
                 if (!flusher->try_find_other(dirty_end, cur))
                 {
+                    cur.oid = {};
                     goto stop_flusher;
                 }
             }
         }
 #ifdef BLOCKSTORE_DEBUG
-        printf("Flushing %lx:%lx v%lu\n", cur.oid.inode, cur.oid.stripe, cur.version);
+        printf("Flushing %jx:%jx v%ju\n", cur.oid.inode, cur.oid.stripe, cur.version);
 #endif
         flusher->active_flushers++;
         // Find it in clean_db
@@ -448,7 +448,7 @@ stop_flusher:
                 // Object not allocated. This is a bug.
                 char err[1024];
                 snprintf(
-                    err, 1024, "BUG: Object %lx:%lx v%lu that we are trying to flush is not allocated on the data device",
+                    err, 1024, "BUG: Object %jx:%jx v%ju that we are trying to flush is not allocated on the data device",
                     cur.oid.inode, cur.oid.stripe, cur.version
                 );
                 throw std::runtime_error(err);
@@ -538,7 +538,7 @@ resume_2:
                 clean_disk_entry *old_entry = (clean_disk_entry*)((uint8_t*)meta_old.buf + meta_old.pos*bs->dsk.clean_entry_size);
                 if (old_entry->oid.inode != 0 && old_entry->oid != cur.oid)
                 {
-                    printf("Fatal error (metadata corruption or bug): tried to wipe metadata entry %lu (%lx:%lx v%lu) as old location of %lx:%lx\n",
+                    printf("Fatal error (metadata corruption or bug): tried to wipe metadata entry %ju (%jx:%jx v%ju) as old location of %jx:%jx\n",
                         old_clean_loc >> bs->dsk.block_order, old_entry->oid.inode, old_entry->oid.stripe,
                         old_entry->version, cur.oid.inode, cur.oid.stripe);
                     exit(1);
@@ -571,7 +571,7 @@ resume_2:
         // Erase dirty_db entries
         bs->erase_dirty(dirty_start, std::next(dirty_end), clean_loc);
 #ifdef BLOCKSTORE_DEBUG
-        printf("Flushed %lx:%lx v%lu (%d copies, wr:%d, del:%d), %ld left\n", cur.oid.inode, cur.oid.stripe, cur.version,
+        printf("Flushed %jx:%jx v%ju (%d copies, wr:%d, del:%d), %jd left\n", cur.oid.inode, cur.oid.stripe, cur.version,
             copy_count, has_writes, has_delete, flusher->flush_queue.size());
 #endif
     release_oid:
@@ -584,7 +584,8 @@ resume_2:
         flusher->sync_to_repeat.erase(repeat_it);
     trim_journal:
         // Clear unused part of the journal every <journal_trim_interval> flushes
-        if (!((++flusher->journal_trim_counter) % flusher->journal_trim_interval) || flusher->trim_wanted > 0)
+        if (bs->journal_trim_interval && !((++flusher->journal_trim_counter) % bs->journal_trim_interval) ||
+            flusher->trim_wanted > 0)
         {
     resume_26:
     resume_27:
@@ -609,8 +610,8 @@ void journal_flusher_co::update_metadata_entry()
     {
         printf(
             has_delete
-                ? "Fatal error (metadata corruption or bug): tried to delete metadata entry %lu (%lx:%lx v%lu) while deleting %lx:%lx v%lu\n"
+                ? "Fatal error (metadata corruption or bug): tried to delete metadata entry %ju (%jx:%jx v%ju) while deleting %jx:%jx v%ju\n"
-                : "Fatal error (metadata corruption or bug): tried to overwrite non-zero metadata entry %lu (%lx:%lx v%lu) with %lx:%lx v%lu\n",
+                : "Fatal error (metadata corruption or bug): tried to overwrite non-zero metadata entry %ju (%jx:%jx v%ju) with %jx:%jx v%ju\n",
             clean_loc >> bs->dsk.block_order, new_entry->oid.inode, new_entry->oid.stripe,
             new_entry->version, cur.oid.inode, cur.oid.stripe, cur.version
         );
@@ -710,7 +711,7 @@ bool journal_flusher_co::write_meta_block(flusher_meta_write_t & meta_block, int
     if (wait_state == wait_base)
         goto resume_0;
     await_sqe(0);
-    data->iov = (struct iovec){ meta_block.buf, bs->dsk.meta_block_size };
+    data->iov = (struct iovec){ meta_block.buf, (size_t)bs->dsk.meta_block_size };
     data->callback = simple_callback_w;
     my_uring_prep_writev(
         sqe, bs->dsk.meta_fd, &data->iov, 1, bs->dsk.meta_offset + bs->dsk.meta_block_size + meta_block.sector
@@ -760,7 +761,7 @@ bool journal_flusher_co::clear_incomplete_csum_block_bits(int wait_base)
         {
             // If we encounter bad checksums during flush, we still update the bad block,
             // but intentionally mangle checksums to avoid hiding the corruption.
-            iovec iov = { .iov_base = v[i].buf, .iov_len = v[i].len };
+            iovec iov = { .iov_base = v[i].buf, .iov_len = (size_t)v[i].len };
             if (!(v[i].copy_flags & COPY_BUF_JOURNAL))
             {
                 assert(!(v[i].offset % bs->dsk.csum_block_size));
@@ -768,7 +769,7 @@ bool journal_flusher_co::clear_incomplete_csum_block_bits(int wait_base)
                 bs->verify_padded_checksums(new_clean_bitmap, new_clean_bitmap + 2*bs->dsk.clean_entry_bitmap_size,
                     v[i].offset, &iov, 1, [&](uint32_t bad_block, uint32_t calc_csum, uint32_t stored_csum)
                 {
-                    printf("Checksum mismatch in object %lx:%lx v%lu in data area at offset 0x%lx+0x%x: got %08x, expected %08x\n",
+                    printf("Checksum mismatch in object %jx:%jx v%ju in data area at offset 0x%jx+0x%x: got %08x, expected %08x\n",
                         cur.oid.inode, cur.oid.stripe, old_clean_ver, old_clean_loc, bad_block, calc_csum, stored_csum);
                     for (uint32_t j = 0; j < bs->dsk.csum_block_size; j += bs->dsk.bitmap_granularity)
                     {
@@ -781,7 +782,7 @@ bool journal_flusher_co::clear_incomplete_csum_block_bits(int wait_base)
             {
                 bs->verify_journal_checksums(v[i].csum_buf, v[i].offset, &iov, 1, [&](uint32_t bad_block, uint32_t calc_csum, uint32_t stored_csum)
                 {
-                    printf("Checksum mismatch in object %lx:%lx v%lu in journal at offset 0x%lx+0x%x (block offset 0x%lx): got %08x, expected %08x\n",
+                    printf("Checksum mismatch in object %jx:%jx v%ju in journal at offset 0x%jx+0x%x (block offset 0x%jx): got %08x, expected %08x\n",
                         cur.oid.inode, cur.oid.stripe, old_clean_ver,
                         v[i].disk_offset, bad_block, v[i].offset, calc_csum, stored_csum);
                     bad_block += (v[i].offset/bs->dsk.csum_block_size) * bs->dsk.csum_block_size;
@@ -805,7 +806,7 @@ bool journal_flusher_co::clear_incomplete_csum_block_bits(int wait_base)
             if (new_entry->oid != cur.oid)
             {
                 printf(
-                    "Fatal error (metadata corruption or bug): tried to make holes in %lu (%lx:%lx v%lu) with %lx:%lx v%lu\n",
+                    "Fatal error (metadata corruption or bug): tried to make holes in %ju (%jx:%jx v%ju) with %jx:%jx v%ju\n",
                     clean_loc >> bs->dsk.block_order, new_entry->oid.inode, new_entry->oid.stripe,
                     new_entry->version, cur.oid.inode, cur.oid.stripe, cur.version
                 );
@@ -925,7 +926,7 @@ void journal_flusher_co::scan_dirty()
         {
             char err[1024];
             snprintf(
-                err, 1024, "BUG: Unexpected dirty_entry %lx:%lx v%lu unstable state during flush: 0x%x",
+                err, 1024, "BUG: Unexpected dirty_entry %jx:%jx v%ju unstable state during flush: 0x%x",
                 dirty_it->first.oid.inode, dirty_it->first.oid.stripe, dirty_it->first.version, dirty_it->second.state
             );
             throw std::runtime_error(err);
@@ -1021,7 +1022,7 @@ void journal_flusher_co::scan_dirty()
             // May happen if the metadata entry is corrupt, but journal isn't
             // FIXME: Report corrupted object to the upper layer (OSD)
             printf(
-                "Warning: object %lx:%lx has overwrites, but doesn't have a clean version."
+                "Warning: object %jx:%jx has overwrites, but doesn't have a clean version."
                 " Metadata is likely corrupted. Dropping object from the DB.\n",
                 cur.oid.inode, cur.oid.stripe
             );
@@ -1056,7 +1057,7 @@ void journal_flusher_co::scan_dirty()
         flusher->enqueue_flush(cur);
         cur.version = dirty_end->first.version;
 #ifdef BLOCKSTORE_DEBUG
-        printf("Partial checksum block overwrites found - rewinding flush back to %lx:%lx v%lu\n", cur.oid.inode, cur.oid.stripe, cur.version);
+        printf("Partial checksum block overwrites found - rewinding flush back to %jx:%jx v%ju\n", cur.oid.inode, cur.oid.stripe, cur.version);
 #endif
         v.clear();
         copy_count = 0;
@@ -1084,7 +1085,7 @@ bool journal_flusher_co::read_dirty(int wait_base)
         auto & vi = v[v.size()-i];
         assert(vi.len != 0);
         vi.buf = memalign_or_die(MEM_ALIGNMENT, vi.len);
-        data->iov = (struct iovec){ vi.buf, vi.len };
+        data->iov = (struct iovec){ vi.buf, (size_t)vi.len };
         data->callback = simple_callback_r;
         my_uring_prep_readv(
             sqe, bs->dsk.data_fd, &data->iov, 1, bs->dsk.data_offset + old_clean_loc + vi.offset
@@ -1208,7 +1209,7 @@ bool journal_flusher_co::modify_meta_read(uint64_t meta_loc, flusher_meta_write_
             .usage_count = 1,
         }).first;
         await_sqe(0);
-        data->iov = (struct iovec){ wr.it->second.buf, bs->dsk.meta_block_size };
+        data->iov = (struct iovec){ wr.it->second.buf, (size_t)bs->dsk.meta_block_size };
         data->callback = simple_callback_r;
         wr.submitted = true;
         my_uring_prep_readv(
@@ -1247,7 +1248,7 @@ void journal_flusher_co::free_data_blocks()
         auto uo_it = bs->used_clean_objects.find(old_clean_loc);
         bool used = uo_it != bs->used_clean_objects.end();
 #ifdef BLOCKSTORE_DEBUG
-        printf("%s block %lu from %lx:%lx v%lu (new location is %lu)\n",
+        printf("%s block %ju from %jx:%jx v%ju (new location is %ju)\n",
             used ? "Postpone free" : "Free",
             old_clean_loc >> bs->dsk.block_order,
             cur.oid.inode, cur.oid.stripe, cur.version,
@@ -1264,7 +1265,7 @@ void journal_flusher_co::free_data_blocks()
         auto uo_it = bs->used_clean_objects.find(old_clean_loc);
         bool used = uo_it != bs->used_clean_objects.end();
 #ifdef BLOCKSTORE_DEBUG
-        printf("%s block %lu from %lx:%lx v%lu (delete)\n",
+        printf("%s block %ju from %jx:%jx v%ju (delete)\n",
             used ? "Postpone free" : "Free",
             old_clean_loc >> bs->dsk.block_order,
             cur.oid.inode, cur.oid.stripe, cur.version);
@@ -1346,7 +1347,6 @@ bool journal_flusher_co::trim_journal(int wait_base)
     else if (wait_state == wait_base+2) goto resume_2;
     else if (wait_state == wait_base+3) goto resume_3;
     else if (wait_state == wait_base+4) goto resume_4;
-    flusher->journal_trim_counter = 0;
     new_trim_pos = bs->journal.get_trim_pos();
     if (new_trim_pos != bs->journal.used_start)
     {
@@ -1378,7 +1378,7 @@ bool journal_flusher_co::trim_journal(int wait_base)
                 .csum_block_size = bs->dsk.csum_block_size,
             };
             ((journal_entry_start*)flusher->journal_superblock)->crc32 = je_crc32((journal_entry*)flusher->journal_superblock);
-            data->iov = (struct iovec){ flusher->journal_superblock, bs->dsk.journal_block_size };
+            data->iov = (struct iovec){ flusher->journal_superblock, (size_t)bs->dsk.journal_block_size };
             data->callback = simple_callback_w;
             my_uring_prep_writev(sqe, bs->dsk.journal_fd, &data->iov, 1, bs->journal.offset);
             wait_count++;
@@ -1410,7 +1410,7 @@ bool journal_flusher_co::trim_journal(int wait_base)
             }
             bs->journal.used_start = new_trim_pos;
 #ifdef BLOCKSTORE_DEBUG
-            printf("Journal trimmed to %08lx (next_free=%08lx dirty_start=%08lx)\n", bs->journal.used_start, bs->journal.next_free, bs->journal.dirty_start);
+            printf("Journal trimmed to %08jx (next_free=%08jx dirty_start=%08jx)\n", bs->journal.used_start, bs->journal.next_free, bs->journal.dirty_start);
 #endif
             if (bs->journal.flush_journal && !flusher->flush_queue.size())
             {
@@ -1419,6 +1419,7 @@ bool journal_flusher_co::trim_journal(int wait_base)
                 exit(0);
             }
         }
+        flusher->journal_trim_counter = 0;
         flusher->trimming = false;
     }
     return true;

src/blockstore_flush.h

@@ -107,7 +107,7 @@ class journal_flusher_t
     blockstore_impl_t *bs;
     friend class journal_flusher_co;
 
-    int journal_trim_counter, journal_trim_interval;
+    int journal_trim_counter;
     bool trimming;
     void* journal_superblock;
 

src/blockstore_impl.cpp

@@ -195,6 +195,10 @@ void blockstore_impl_t::loop()
                     // ring is full, stop submission
                     break;
                 }
+                else if (PRIV(op)->wait_for == WAIT_JOURNAL)
+                {
+                    PRIV(op)->wait_detail2 = (unstable_writes.size()+unstable_unsynced);
+                }
             }
         }
         if (op_idx != new_idx)
@@ -265,7 +269,7 @@ void blockstore_impl_t::check_wait(blockstore_op_t *op)
         {
             // stop submission if there's still no free space
 #ifdef BLOCKSTORE_DEBUG
-            printf("Still waiting for %lu SQE(s)\n", PRIV(op)->wait_detail);
+            printf("Still waiting for %ju SQE(s)\n", PRIV(op)->wait_detail);
 #endif
             return;
         }
@@ -273,15 +277,15 @@ void blockstore_impl_t::check_wait(blockstore_op_t *op)
     }
     else if (PRIV(op)->wait_for == WAIT_JOURNAL)
     {
-        if (journal.used_start == PRIV(op)->wait_detail && !unstable_count_changed)
+        if (journal.used_start == PRIV(op)->wait_detail &&
+            (unstable_writes.size()+unstable_unsynced) == PRIV(op)->wait_detail2)
         {
             // do not submit
 #ifdef BLOCKSTORE_DEBUG
-            printf("Still waiting to flush journal offset %08lx\n", PRIV(op)->wait_detail);
+            printf("Still waiting to flush journal offset %08jx\n", PRIV(op)->wait_detail);
 #endif
             return;
         }
-        unstable_count_changed = false;
         flusher->release_trim();
         PRIV(op)->wait_for = 0;
     }
@@ -353,7 +357,6 @@ void blockstore_impl_t::enqueue_op(blockstore_op_t *op)
                     };
                 }
                 unstable_writes.clear();
-                unstable_count_changed = true;
                 op->callback = [old_callback](blockstore_op_t *op)
                 {
                     obj_ver_id *vers = (obj_ver_id*)op->buf;

src/blockstore_impl.h

@@ -202,7 +202,7 @@ struct blockstore_op_private_t
 {
     // Wait status
     int wait_for;
-    uint64_t wait_detail;
+    uint64_t wait_detail, wait_detail2;
     int pending_ops;
     int op_state;
 
@@ -253,6 +253,7 @@ class blockstore_impl_t
     bool inmemory_meta = false;
     // Maximum and minimum flusher count
     unsigned max_flusher_count, min_flusher_count;
+    unsigned journal_trim_interval;
     // Maximum queue depth
     unsigned max_write_iodepth = 128;
     // Enable small (journaled) write throttling, useful for the SSD+HDD case
@@ -276,7 +277,6 @@ class blockstore_impl_t
     std::vector<blockstore_op_t*> submit_queue;
     std::vector<obj_ver_id> unsynced_big_writes, unsynced_small_writes;
     int unsynced_big_write_count = 0, unstable_unsynced = 0;
-    bool unstable_count_changed = false;
     int unsynced_queued_ops = 0;
     allocator *data_alloc = NULL;
     uint64_t used_blocks = 0;

src/blockstore_init.cpp

@@ -63,7 +63,7 @@ int blockstore_init_meta::loop()
         throw std::runtime_error("Failed to allocate metadata read buffer");
     // Read superblock
     GET_SQE();
-    data->iov = { metadata_buffer, bs->dsk.meta_block_size };
+    data->iov = { metadata_buffer, (size_t)bs->dsk.meta_block_size };
     data->callback = [this](ring_data_t *data) { handle_event(data, -1); };
     my_uring_prep_readv(sqe, bs->dsk.meta_fd, &data->iov, 1, bs->dsk.meta_offset);
     bs->ringloop->submit();
@@ -100,7 +100,7 @@ resume_1:
         {
             printf("Initializing metadata area\n");
             GET_SQE();
-            data->iov = (struct iovec){ metadata_buffer, bs->dsk.meta_block_size };
+            data->iov = (struct iovec){ metadata_buffer, (size_t)bs->dsk.meta_block_size };
             data->callback = [this](ring_data_t *data) { handle_event(data, -1); };
             my_uring_prep_writev(sqe, bs->dsk.meta_fd, &data->iov, 1, bs->dsk.meta_offset);
             bs->ringloop->submit();
@@ -153,7 +153,7 @@ resume_1:
         else if (hdr->version > BLOCKSTORE_META_FORMAT_V2)
         {
             printf(
-                "Metadata format is too new for me (stored version is %lu, max supported %u).\n",
+                "Metadata format is too new for me (stored version is %ju, max supported %u).\n",
                 hdr->version, BLOCKSTORE_META_FORMAT_V2
             );
             exit(1);
@@ -167,7 +167,7 @@ resume_1:
             printf(
                 "Configuration stored in metadata superblock"
                 " (meta_block_size=%u, data_block_size=%u, bitmap_granularity=%u, data_csum_type=%u, csum_block_size=%u)"
-                " differs from OSD configuration (%lu/%u/%lu, %u/%u).\n",
+                " differs from OSD configuration (%ju/%u/%ju, %u/%u).\n",
                 hdr->meta_block_size, hdr->data_block_size, hdr->bitmap_granularity,
                 hdr->data_csum_type, hdr->csum_block_size,
                 bs->dsk.meta_block_size, bs->dsk.data_block_size, bs->dsk.bitmap_granularity,
@@ -199,7 +199,8 @@ resume_2:
                 submitted++;
                 next_offset += bufs[i].size;
                 GET_SQE();
-                data->iov = { bufs[i].buf, bufs[i].size };
+                assert(bufs[i].size <= 0x7fffffff);
+                data->iov = { bufs[i].buf, (size_t)bufs[i].size };
                 data->callback = [this, i](ring_data_t *data) { handle_event(data, i); };
                 if (!zero_on_init)
                     my_uring_prep_readv(sqe, bs->dsk.meta_fd, &data->iov, 1, bs->dsk.meta_offset + bufs[i].offset);
@@ -231,7 +232,8 @@ resume_2:
             {
                 // write the modified buffer back
                 GET_SQE();
-                data->iov = { bufs[i].buf, bufs[i].size };
+                assert(bufs[i].size <= 0x7fffffff);
+                data->iov = { bufs[i].buf, (size_t)bufs[i].size };
                 data->callback = [this, i](ring_data_t *data) { handle_event(data, i); };
                 my_uring_prep_writev(sqe, bs->dsk.meta_fd, &data->iov, 1, bs->dsk.meta_offset + bufs[i].offset);
                 bufs[i].state = INIT_META_WRITING;
@@ -257,7 +259,7 @@ resume_2:
             next_offset = entries_to_zero[i]/entries_per_block;
             for (j = i; j < entries_to_zero.size() && entries_to_zero[j]/entries_per_block == next_offset; j++) {}
             GET_SQE();
-            data->iov = { metadata_buffer, bs->dsk.meta_block_size };
+            data->iov = { metadata_buffer, (size_t)bs->dsk.meta_block_size };
             data->callback = [this](ring_data_t *data) { handle_event(data, -1); };
             my_uring_prep_readv(sqe, bs->dsk.meta_fd, &data->iov, 1, bs->dsk.meta_offset + (1+next_offset)*bs->dsk.meta_block_size);
             submitted++;
@@ -273,7 +275,7 @@ resume_5:
                 memset((uint8_t*)metadata_buffer + pos*bs->dsk.clean_entry_size, 0, bs->dsk.clean_entry_size);
             }
             GET_SQE();
-            data->iov = { metadata_buffer, bs->dsk.meta_block_size };
+            data->iov = { metadata_buffer, (size_t)bs->dsk.meta_block_size };
             data->callback = [this](ring_data_t *data) { handle_event(data, -1); };
             my_uring_prep_writev(sqe, bs->dsk.meta_fd, &data->iov, 1, bs->dsk.meta_offset + (1+next_offset)*bs->dsk.meta_block_size);
             submitted++;
@@ -287,7 +289,7 @@ resume_6:
         entries_to_zero.clear();
     }
     // metadata read finished
-    printf("Metadata entries loaded: %lu, free blocks: %lu / %lu\n", entries_loaded, bs->data_alloc->get_free_count(), bs->dsk.block_count);
+    printf("Metadata entries loaded: %ju, free blocks: %ju / %ju\n", entries_loaded, bs->data_alloc->get_free_count(), bs->dsk.block_count);
     if (!bs->inmemory_meta)
     {
         free(metadata_buffer);
@@ -328,7 +330,7 @@ bool blockstore_init_meta::handle_meta_block(uint8_t *buf, uint64_t entries_per_
                 uint32_t *entry_csum = (uint32_t*)((uint8_t*)entry + bs->dsk.clean_entry_size - 4);
                 if (*entry_csum != crc32c(0, entry, bs->dsk.clean_entry_size - 4))
                 {
-                    printf("Metadata entry %lu is corrupt (checksum mismatch), skipping\n", done_cnt+i);
+                    printf("Metadata entry %ju is corrupt (checksum mismatch), skipping\n", done_cnt+i);
                     continue;
                 }
             }
@@ -366,7 +368,7 @@ bool blockstore_init_meta::handle_meta_block(uint8_t *buf, uint64_t entries_per_
                         entries_to_zero.push_back(clean_it->second.location >> bs->dsk.block_order);
                     }
 #ifdef BLOCKSTORE_DEBUG
-                    printf("Free block %lu from %lx:%lx v%lu (new location is %lu)\n",
+                    printf("Free block %ju from %jx:%jx v%ju (new location is %ju)\n",
                         old_clean_loc,
                         clean_it->first.inode, clean_it->first.stripe, clean_it->second.version,
                         done_cnt+i);
@@ -380,7 +382,7 @@ bool blockstore_init_meta::handle_meta_block(uint8_t *buf, uint64_t entries_per_
                 }
                 entries_loaded++;
 #ifdef BLOCKSTORE_DEBUG
-                printf("Allocate block (clean entry) %lu: %lx:%lx v%lu\n", done_cnt+i, entry->oid.inode, entry->oid.stripe, entry->version);
+                printf("Allocate block (clean entry) %ju: %jx:%jx v%ju\n", done_cnt+i, entry->oid.inode, entry->oid.stripe, entry->version);
 #endif
                 bs->data_alloc->set(done_cnt+i, true);
                 clean_db[entry->oid] = (struct clean_entry){
@@ -394,7 +396,7 @@ bool blockstore_init_meta::handle_meta_block(uint8_t *buf, uint64_t entries_per_
                 updated = true;
                 memset(entry, 0, bs->dsk.clean_entry_size);
 #ifdef BLOCKSTORE_DEBUG
-                printf("Old clean entry %lu: %lx:%lx v%lu\n", done_cnt+i, entry->oid.inode, entry->oid.stripe, entry->version);
+                printf("Old clean entry %ju: %jx:%jx v%ju\n", done_cnt+i, entry->oid.inode, entry->oid.stripe, entry->version);
 #endif
             }
         }
@@ -466,7 +468,7 @@ int blockstore_init_journal::loop()
     if (!sqe)
         throw std::runtime_error("io_uring is full while trying to read journal");
     data = ((ring_data_t*)sqe->user_data);
-    data->iov = { submitted_buf, bs->journal.block_size };
+    data->iov = { submitted_buf, (size_t)bs->journal.block_size };
     data->callback = simple_callback;
     my_uring_prep_readv(sqe, bs->dsk.journal_fd, &data->iov, 1, bs->journal.offset);
     bs->ringloop->submit();
@@ -507,7 +509,7 @@ resume_1:
             // FIXME: Randomize initial crc32. Track crc32 when trimming.
             printf("Resetting journal\n");
             GET_SQE();
-            data->iov = (struct iovec){ submitted_buf, 2*bs->journal.block_size };
+            data->iov = (struct iovec){ submitted_buf, (size_t)(2*bs->journal.block_size) };
             data->callback = simple_callback;
             my_uring_prep_writev(sqe, bs->dsk.journal_fd, &data->iov, 1, bs->journal.offset);
             wait_count++;
@@ -557,7 +559,7 @@ resume_1:
             (je_start->version != JOURNAL_VERSION_V2 || je_start->size != JE_START_V2_SIZE && je_start->size != JE_START_V1_SIZE))
         {
             fprintf(
-                stderr, "The code only supports journal versions 2 and 1, but it is %lu on disk."
+                stderr, "The code only supports journal versions 2 and 1, but it is %ju on disk."
                     " Please use vitastor-disk to rewrite the journal\n",
                 je_start->size == JE_START_V0_SIZE ? 0 : je_start->version
             );
@@ -606,7 +608,7 @@ resume_1:
                     submitted_buf = (uint8_t*)bs->journal.buffer + journal_pos;
                 data->iov = {
                     submitted_buf,
-                    end - journal_pos < JOURNAL_BUFFER_SIZE ? end - journal_pos : JOURNAL_BUFFER_SIZE,
+                    (size_t)(end - journal_pos < JOURNAL_BUFFER_SIZE ? end - journal_pos : JOURNAL_BUFFER_SIZE),
                 };
                 data->callback = [this](ring_data_t *data1) { handle_event(data1); };
                 my_uring_prep_readv(sqe, bs->dsk.journal_fd, &data->iov, 1, bs->journal.offset + journal_pos);
@@ -622,7 +624,7 @@ resume_1:
                     if (init_write_buf && !bs->readonly)
                     {
                         GET_SQE();
-                        data->iov = { init_write_buf, bs->journal.block_size };
+                        data->iov = { init_write_buf, (size_t)bs->journal.block_size };
                         data->callback = simple_callback;
                         my_uring_prep_writev(sqe, bs->dsk.journal_fd, &data->iov, 1, bs->journal.offset + init_write_sector);
                         wait_count++;
@@ -691,7 +693,7 @@ resume_1:
             IS_BIG_WRITE(dirty_it->second.state) &&
             dirty_it->second.location == UINT64_MAX)
         {
-            printf("Fatal error (bug): %lx:%lx v%lu big_write journal_entry was allocated over another object\n",
+            printf("Fatal error (bug): %jx:%jx v%ju big_write journal_entry was allocated over another object\n",
                 dirty_it->first.oid.inode, dirty_it->first.oid.stripe, dirty_it->first.version);
             exit(1);
         }
@@ -699,7 +701,7 @@ resume_1:
     bs->flusher->mark_trim_possible();
     bs->journal.dirty_start = bs->journal.next_free;
     printf(
-        "Journal entries loaded: %lu, free journal space: %lu bytes (%08lx..%08lx is used), free blocks: %lu / %lu\n",
+        "Journal entries loaded: %ju, free journal space: %ju bytes (%08jx..%08jx is used), free blocks: %ju / %ju\n",
         entries_loaded,
         (bs->journal.next_free >= bs->journal.used_start
             ? bs->journal.len-bs->journal.block_size - (bs->journal.next_free-bs->journal.used_start)
@@ -754,7 +756,7 @@ int blockstore_init_journal::handle_journal_part(void *buf, uint64_t done_pos, u
             {
 #ifdef BLOCKSTORE_DEBUG
                 printf(
-                    "je_small_write%s oid=%lx:%lx ver=%lu offset=%u len=%u\n",
+                    "je_small_write%s oid=%jx:%jx ver=%ju offset=%u len=%u\n",
                     je->type == JE_SMALL_WRITE_INSTANT ? "_instant" : "",
                     je->small_write.oid.inode, je->small_write.oid.stripe, je->small_write.version,
                     je->small_write.offset, je->small_write.len
@@ -776,7 +778,7 @@ int blockstore_init_journal::handle_journal_part(void *buf, uint64_t done_pos, u
                 if (location != je->small_write.data_offset)
                 {
                     char err[1024];
-                    snprintf(err, 1024, "BUG: calculated journal data offset (%08lx) != stored journal data offset (%08lx)", location, je->small_write.data_offset);
+                    snprintf(err, 1024, "BUG: calculated journal data offset (%08jx) != stored journal data offset (%08jx)", location, je->small_write.data_offset);
                     throw std::runtime_error(err);
                 }
                 small_write_data.clear();
@@ -803,7 +805,7 @@ int blockstore_init_journal::handle_journal_part(void *buf, uint64_t done_pos, u
                             covered += part_end - part_begin;
                             small_write_data.push_back((iovec){
                                 .iov_base = (uint8_t*)done[i].buf + part_begin - done[i].pos,
-                                .iov_len = part_end - part_begin,
+                                .iov_len = (size_t)(part_end - part_begin),
                             });
                         }
                     }
@@ -826,7 +828,7 @@ int blockstore_init_journal::handle_journal_part(void *buf, uint64_t done_pos, u
                     if (!data_csum_valid)
                     {
                         printf(
-                            "Journal entry data is corrupt for small_write%s oid=%lx:%lx ver=%lu offset=%u len=%u - data crc32 %x != %x\n",
+                            "Journal entry data is corrupt for small_write%s oid=%jx:%jx ver=%ju offset=%u len=%u - data crc32 %x != %x\n",
                             je->type == JE_SMALL_WRITE_INSTANT ? "_instant" : "",
                             je->small_write.oid.inode, je->small_write.oid.stripe, je->small_write.version,
                             je->small_write.offset, je->small_write.len,
@@ -845,7 +847,7 @@ int blockstore_init_journal::handle_journal_part(void *buf, uint64_t done_pos, u
                     if (je->size != required_size)
                     {
                         printf(
-                            "Journal entry data has invalid size for small_write%s oid=%lx:%lx ver=%lu offset=%u len=%u - should be %u bytes but is %u bytes\n",
+                            "Journal entry data has invalid size for small_write%s oid=%jx:%jx ver=%ju offset=%u len=%u - should be %u bytes but is %u bytes\n",
                             je->type == JE_SMALL_WRITE_INSTANT ? "_instant" : "",
                             je->small_write.oid.inode, je->small_write.oid.stripe, je->small_write.version,
                             je->small_write.offset, je->small_write.len,
@@ -893,7 +895,7 @@ int blockstore_init_journal::handle_journal_part(void *buf, uint64_t done_pos, u
                             if (block_crc32 != *block_csums)
                             {
                                 printf(
-                                    "Journal entry data is corrupt for small_write%s oid=%lx:%lx ver=%lu offset=%u len=%u - block %u crc32 %x != %x\n",
+                                    "Journal entry data is corrupt for small_write%s oid=%jx:%jx ver=%ju offset=%u len=%u - block %u crc32 %x != %x\n",
                                     je->type == JE_SMALL_WRITE_INSTANT ? "_instant" : "",
                                     je->small_write.oid.inode, je->small_write.oid.stripe, je->small_write.version,
                                     je->small_write.offset, je->small_write.len,
@@ -956,7 +958,7 @@ int blockstore_init_journal::handle_journal_part(void *buf, uint64_t done_pos, u
                     bs->journal.used_sectors[proc_pos]++;
 #ifdef BLOCKSTORE_DEBUG
                     printf(
-                        "journal offset %08lx is used by %lx:%lx v%lu (%lu refs)\n",
+                        "journal offset %08jx is used by %jx:%jx v%ju (%ju refs)\n",
                         proc_pos, ov.oid.inode, ov.oid.stripe, ov.version, bs->journal.used_sectors[proc_pos]
                     );
 #endif
@@ -972,7 +974,7 @@ int blockstore_init_journal::handle_journal_part(void *buf, uint64_t done_pos, u
             {
 #ifdef BLOCKSTORE_DEBUG
                 printf(
-                    "je_big_write%s oid=%lx:%lx ver=%lu loc=%lu\n",
+                    "je_big_write%s oid=%jx:%jx ver=%ju loc=%ju\n",
                     je->type == JE_BIG_WRITE_INSTANT ? "_instant" : "",
                     je->big_write.oid.inode, je->big_write.oid.stripe, je->big_write.version, je->big_write.location >> bs->dsk.block_order
                 );
@@ -1049,7 +1051,7 @@ int blockstore_init_journal::handle_journal_part(void *buf, uint64_t done_pos, u
                     {
 #ifdef BLOCKSTORE_DEBUG
                         printf(
-                            "Allocate block (journal) %lu: %lx:%lx v%lu\n",
+                            "Allocate block (journal) %ju: %jx:%jx v%ju\n",
                             je->big_write.location >> bs->dsk.block_order,
                             ov.oid.inode, ov.oid.stripe, ov.version
                         );
@@ -1059,7 +1061,7 @@ int blockstore_init_journal::handle_journal_part(void *buf, uint64_t done_pos, u
                     bs->journal.used_sectors[proc_pos]++;
 #ifdef BLOCKSTORE_DEBUG
                     printf(
-                        "journal offset %08lx is used by %lx:%lx v%lu (%lu refs)\n",
+                        "journal offset %08jx is used by %jx:%jx v%ju (%ju refs)\n",
                         proc_pos, ov.oid.inode, ov.oid.stripe, ov.version, bs->journal.used_sectors[proc_pos]
                     );
 #endif
@@ -1074,7 +1076,7 @@ int blockstore_init_journal::handle_journal_part(void *buf, uint64_t done_pos, u
             else if (je->type == JE_STABLE)
             {
 #ifdef BLOCKSTORE_DEBUG
-                printf("je_stable oid=%lx:%lx ver=%lu\n", je->stable.oid.inode, je->stable.oid.stripe, je->stable.version);
+                printf("je_stable oid=%jx:%jx ver=%ju\n", je->stable.oid.inode, je->stable.oid.stripe, je->stable.version);
 #endif
                 // oid, version
                 obj_ver_id ov = {
@@ -1086,7 +1088,7 @@ int blockstore_init_journal::handle_journal_part(void *buf, uint64_t done_pos, u
             else if (je->type == JE_ROLLBACK)
             {
 #ifdef BLOCKSTORE_DEBUG
-                printf("je_rollback oid=%lx:%lx ver=%lu\n", je->rollback.oid.inode, je->rollback.oid.stripe, je->rollback.version);
+                printf("je_rollback oid=%jx:%jx ver=%ju\n", je->rollback.oid.inode, je->rollback.oid.stripe, je->rollback.version);
 #endif
                 // rollback dirty writes of <oid> up to <version>
                 obj_ver_id ov = {
@@ -1098,7 +1100,7 @@ int blockstore_init_journal::handle_journal_part(void *buf, uint64_t done_pos, u
             else if (je->type == JE_DELETE)
             {
 #ifdef BLOCKSTORE_DEBUG
-                printf("je_delete oid=%lx:%lx ver=%lu\n", je->del.oid.inode, je->del.oid.stripe, je->del.version);
+                printf("je_delete oid=%jx:%jx ver=%ju\n", je->del.oid.inode, je->del.oid.stripe, je->del.version);
 #endif
                 bool dirty_exists = false;
                 auto dirty_it = bs->dirty_db.upper_bound((obj_ver_id){

src/blockstore_journal.cpp

@@ -90,8 +90,8 @@ int blockstore_journal_check_t::check_available(blockstore_op_t *op, int entries
             }
             // In fact, it's even more rare than "ran out of journal space", so print a warning
             printf(
-                "Ran out of journal sector buffers: %d/%lu buffers used (%d dirty), next buffer (%ld)"
+                "Ran out of journal sector buffers: %d/%ju buffers used (%d dirty), next buffer (%jd)"
-                " is %s and flushed %lu times. Consider increasing \'journal_sector_buffer_count\'\n",
+                " is %s and flushed %ju times. Consider increasing \'journal_sector_buffer_count\'\n",
                 used, bs->journal.sector_count, dirty, next_sector,
                 bs->journal.sector_info[next_sector].dirty ? "dirty" : "not dirty",
                 bs->journal.sector_info[next_sector].flush_count
@@ -103,7 +103,7 @@ int blockstore_journal_check_t::check_available(blockstore_op_t *op, int entries
     if (data_after > 0)
     {
         next_pos = next_pos + data_after;
-        if (next_pos > bs->journal.len)
+        if (next_pos >= bs->journal.len)
         {
             if (right_dir)
                 next_pos = bs->journal.block_size + data_after;
@@ -114,7 +114,7 @@ int blockstore_journal_check_t::check_available(blockstore_op_t *op, int entries
     {
         // No space in the journal. Wait until used_start changes.
         printf(
-            "Ran out of journal space (used_start=%08lx, next_free=%08lx, dirty_start=%08lx)\n",
+            "Ran out of journal space (used_start=%08jx, next_free=%08jx, dirty_start=%08jx)\n",
             bs->journal.used_start, bs->journal.next_free, bs->journal.dirty_start
         );
         PRIV(op)->wait_for = WAIT_JOURNAL;
@@ -146,7 +146,7 @@ journal_entry* prefill_single_journal_entry(journal_t & journal, uint16_t type,
         journal.in_sector_pos = 0;
         auto next_next_free = (journal.next_free+journal.block_size) < journal.len ? journal.next_free + journal.block_size : journal.block_size;
         // double check that next_free doesn't cross used_start from the left
-        assert(journal.next_free >= journal.used_start || next_next_free < journal.used_start);
+        assert(journal.next_free >= journal.used_start && next_next_free >= journal.next_free || next_next_free < journal.used_start);
         journal.next_free = next_next_free;
         memset(journal.inmemory
             ? (uint8_t*)journal.buffer + journal.sector_info[journal.cur_sector].offset
@@ -183,7 +183,7 @@ void blockstore_impl_t::prepare_journal_sector_write(int cur_sector, blockstore_
             (journal.inmemory
                 ? (uint8_t*)journal.buffer + journal.sector_info[cur_sector].offset
                 : (uint8_t*)journal.sector_buf + journal.block_size*cur_sector),
-            journal.block_size
+            (size_t)journal.block_size
         };
         data->callback = [this, flush_id = journal.submit_id](ring_data_t *data) { handle_journal_write(data, flush_id); };
         my_uring_prep_writev(
@@ -263,7 +263,7 @@ uint64_t journal_t::get_trim_pos()
             // next_free does not need updating during trim
 #ifdef BLOCKSTORE_DEBUG
             printf(
-                "Trimming journal (used_start=%08lx, next_free=%08lx, dirty_start=%08lx, new_start=%08lx, new_refcount=%ld)\n",
+                "Trimming journal (used_start=%08jx, next_free=%08jx, dirty_start=%08jx, new_start=%08jx, new_refcount=%jd)\n",
                 used_start, next_free, dirty_start,
                 journal_used_it->first, journal_used_it->second
             );
@@ -276,7 +276,7 @@ uint64_t journal_t::get_trim_pos()
         // Journal is cleared up to <journal_used_it>
 #ifdef BLOCKSTORE_DEBUG
         printf(
-            "Trimming journal (used_start=%08lx, next_free=%08lx, dirty_start=%08lx, new_start=%08lx, new_refcount=%ld)\n",
+            "Trimming journal (used_start=%08jx, next_free=%08jx, dirty_start=%08jx, new_start=%08jx, new_refcount=%jd)\n",
             used_start, next_free, dirty_start,
             journal_used_it->first, journal_used_it->second
         );
@@ -296,7 +296,7 @@ void journal_t::dump_diagnostics()
         journal_used_it = used_sectors.begin();
     }
     printf(
-        "Journal: used_start=%08lx next_free=%08lx dirty_start=%08lx trim_to=%08lx trim_to_refs=%ld\n",
+        "Journal: used_start=%08jx next_free=%08jx dirty_start=%08jx trim_to=%08jx trim_to_refs=%jd\n",
         used_start, next_free, dirty_start,
         journal_used_it == used_sectors.end() ? 0 : journal_used_it->first,
         journal_used_it == used_sectors.end() ? 0 : journal_used_it->second

src/blockstore_open.cpp

@@ -13,6 +13,7 @@ void blockstore_impl_t::parse_config(blockstore_config_t & config, bool init)
         max_flusher_count = strtoull(config["flusher_count"].c_str(), NULL, 10);
     }
     min_flusher_count = strtoull(config["min_flusher_count"].c_str(), NULL, 10);
+    journal_trim_interval = strtoull(config["journal_trim_interval"].c_str(), NULL, 10);
     max_write_iodepth = strtoull(config["max_write_iodepth"].c_str(), NULL, 10);
     throttle_small_writes = config["throttle_small_writes"] == "true" || config["throttle_small_writes"] == "1" || config["throttle_small_writes"] == "yes";
     throttle_target_iops = strtoull(config["throttle_target_iops"].c_str(), NULL, 10);
@@ -31,6 +32,10 @@ void blockstore_impl_t::parse_config(blockstore_config_t & config, bool init)
     {
         min_flusher_count = 1;
     }
+    if (!journal_trim_interval)
+    {
+        journal_trim_interval = 512;
+    }
     if (!max_write_iodepth)
     {
         max_write_iodepth = 128;

src/blockstore_read.cpp

@@ -25,7 +25,7 @@ int blockstore_impl_t::fulfill_read_push(blockstore_op_t *op, void *buf, uint64_
         return 1;
     }
     BS_SUBMIT_GET_SQE(sqe, data);
-    data->iov = (struct iovec){ buf, len };
+    data->iov = (struct iovec){ buf, (size_t)len };
     PRIV(op)->pending_ops++;
     my_uring_prep_readv(
         sqe,
@@ -505,7 +505,7 @@ int blockstore_impl_t::dequeue_read(blockstore_op_t *read_op)
         for (auto & rv: PRIV(read_op)->read_vec)
         {
             if (rv.journal_sector)
-                journal.used_sectors[rv.journal_sector-1]++;
+                journal.used_sectors.at(rv.journal_sector-1)++;
         }
     }
     read_op->retval = 0;
@@ -700,7 +700,7 @@ uint8_t* blockstore_impl_t::read_clean_meta_block(blockstore_op_t *op, uint64_t
         .buf = buf,
     });
     BS_SUBMIT_GET_SQE(sqe, data);
-    data->iov = (struct iovec){ buf, dsk.meta_block_size };
+    data->iov = (struct iovec){ buf, (size_t)dsk.meta_block_size };
     PRIV(op)->pending_ops++;
     my_uring_prep_readv(sqe, dsk.meta_fd, &data->iov, 1, dsk.meta_offset + dsk.meta_block_size + sector);
     data->callback = [this, op](ring_data_t *data) { handle_read_event(data, op); };
@@ -855,7 +855,7 @@ void blockstore_impl_t::handle_read_event(ring_data_t *data, blockstore_op_t *op
                             {
                                 ok = false;
                                 printf(
-                                    "Checksum mismatch in object %lx:%lx v%lu in journal at 0x%lx, checksum block #%u: got %08x, expected %08x\n",
+                                    "Checksum mismatch in object %jx:%jx v%ju in journal at 0x%jx, checksum block #%u: got %08x, expected %08x\n",
                                     op->oid.inode, op->oid.stripe, op->version,
                                     rv[i].disk_offset, bad_block / dsk.csum_block_size, calc_csum, stored_csum
                                 );
@@ -875,7 +875,7 @@ void blockstore_impl_t::handle_read_event(ring_data_t *data, blockstore_op_t *op
                                 {
                                     ok = false;
                                     printf(
-                                        "Checksum mismatch in object %lx:%lx v%lu in %s data at 0x%lx, checksum block #%u: got %08x, expected %08x\n",
+                                        "Checksum mismatch in object %jx:%jx v%ju in %s data at 0x%jx, checksum block #%u: got %08x, expected %08x\n",
                                         op->oid.inode, op->oid.stripe, op->version,
                                         (rv[i].copy_flags & COPY_BUF_JOURNALED_BIG ? "redirect-write" : "clean"),
                                         rv[i].disk_offset, bad_block / dsk.csum_block_size, calc_csum, stored_csum
@@ -918,7 +918,7 @@ void blockstore_impl_t::handle_read_event(ring_data_t *data, blockstore_op_t *op
                             {
                                 // checksum error
                                 printf(
-                                    "Checksum mismatch in object %lx:%lx v%lu in %s area at offset 0x%lx+0x%lx: %08x vs %08x\n",
+                                    "Checksum mismatch in object %jx:%jx v%ju in %s area at offset 0x%jx+0x%zx: %08x vs %08x\n",
                                     op->oid.inode, op->oid.stripe, op->version,
                                     (vec.copy_flags & COPY_BUF_JOURNAL) ? "journal" : "data", vec.disk_offset, p,
                                     crc32c(0, (uint8_t*)op->buf + vec.offset - op->offset + p, dsk.csum_block_size), *csum
@@ -966,7 +966,7 @@ void blockstore_impl_t::handle_read_event(ring_data_t *data, blockstore_op_t *op
             {
                 if (rv.journal_sector)
                 {
-                    auto used = --journal.used_sectors[rv.journal_sector-1];
+                    auto used = --journal.used_sectors.at(rv.journal_sector-1);
                     if (used == 0)
                     {
                         journal.used_sectors.erase(rv.journal_sector-1);

src/blockstore_rollback.cpp

@@ -162,7 +162,6 @@ void blockstore_impl_t::mark_rolled_back(const obj_ver_id & ov)
                     unstable_writes.erase(unstab_it);
                 else
                     unstab_it->second = max_unstable;
-                unstable_count_changed = true;
             }
         }
     }
@@ -180,7 +179,7 @@ void blockstore_impl_t::erase_dirty(blockstore_dirty_db_t::iterator dirty_start,
     {
         object_id oid = dirty_it->first.oid;
 #ifdef BLOCKSTORE_DEBUG
-        printf("Unblock writes-after-delete %lx:%lx v%lu\n", oid.inode, oid.stripe, dirty_it->first.version);
+        printf("Unblock writes-after-delete %jx:%jx v%ju\n", oid.inode, oid.stripe, dirty_it->first.version);
 #endif
         dirty_it = dirty_end;
         // Unblock operations blocked by delete flushing
@@ -211,21 +210,26 @@ void blockstore_impl_t::erase_dirty(blockstore_dirty_db_t::iterator dirty_start,
             dirty_it->second.location != UINT64_MAX)
         {
 #ifdef BLOCKSTORE_DEBUG
-            printf("Free block %lu from %lx:%lx v%lu\n", dirty_it->second.location >> dsk.block_order,
+            printf("Free block %ju from %jx:%jx v%ju\n", dirty_it->second.location >> dsk.block_order,
                 dirty_it->first.oid.inode, dirty_it->first.oid.stripe, dirty_it->first.version);
 #endif
             data_alloc->set(dirty_it->second.location >> dsk.block_order, false);
         }
-        auto used = --journal.used_sectors[dirty_it->second.journal_sector];
+        auto used = --journal.used_sectors.at(dirty_it->second.journal_sector);
 #ifdef BLOCKSTORE_DEBUG
         printf(
-            "remove usage of journal offset %08lx by %lx:%lx v%lu (%lu refs)\n", dirty_it->second.journal_sector,
+            "remove usage of journal offset %08jx by %jx:%jx v%ju (%ju refs)\n", dirty_it->second.journal_sector,
             dirty_it->first.oid.inode, dirty_it->first.oid.stripe, dirty_it->first.version, used
         );
 #endif
         if (used == 0)
         {
             journal.used_sectors.erase(dirty_it->second.journal_sector);
+            if (dirty_it->second.journal_sector == journal.sector_info[journal.cur_sector].offset)
+            {
+                // Mark current sector as "full" to select the new one
+                journal.in_sector_pos = dsk.journal_block_size;
+            }
             flusher->mark_trim_possible();
         }
         free_dirty_dyn_data(dirty_it->second);

src/blockstore_stable.cpp

@@ -298,7 +298,7 @@ int blockstore_impl_t::dequeue_stable(blockstore_op_t *op)
             if (clean_it == clean_db.end() || clean_it->second.version < ov.version)
             {
                 // No such object version
-                printf("Error: %lx:%lx v%lu not found while stabilizing\n", ov.oid.inode, ov.oid.stripe, ov.version);
+                printf("Error: %jx:%jx v%ju not found while stabilizing\n", ov.oid.inode, ov.oid.stripe, ov.version);
                 return -ENOENT;
             }
             else
@@ -307,35 +307,49 @@ int blockstore_impl_t::dequeue_stable(blockstore_op_t *op)
                 return STAB_SPLIT_DONE;
             }
         }
-        else if (IS_IN_FLIGHT(dirty_it->second.state))
-        {
-            // Object write is still in progress. Wait until the write request completes
-            return STAB_SPLIT_WAIT;
-        }
-        else if (!IS_SYNCED(dirty_it->second.state))
-        {
-            // Object not synced yet - sync it
-            // In previous versions we returned EBUSY here and required
-            // the caller (OSD) to issue a global sync first. But a global sync
-            // waits for all writes in the queue including inflight writes. And
-            // inflight writes may themselves be blocked by unstable writes being
-            // still present in the journal and not flushed away from it.
-            // So we must sync specific objects here.
-            //
-            // Even more, we have to process "stabilize" request in parts. That is,
-            // we must stabilize all objects which are already synced. Otherwise
-            // they may block objects which are NOT synced yet.
-            return STAB_SPLIT_SYNC;
-        }
         else if (IS_STABLE(dirty_it->second.state))
         {
             // Already stable
             return STAB_SPLIT_DONE;
         }
-        else
+        while (true)
         {
-            return STAB_SPLIT_TODO;
+            if (IS_IN_FLIGHT(dirty_it->second.state))
+            {
+                // Object write is still in progress. Wait until the write request completes
+                return STAB_SPLIT_WAIT;
+            }
+            else if (!IS_SYNCED(dirty_it->second.state))
+            {
+                // Object not synced yet - sync it
+                // In previous versions we returned EBUSY here and required
+                // the caller (OSD) to issue a global sync first. But a global sync
+                // waits for all writes in the queue including inflight writes. And
+                // inflight writes may themselves be blocked by unstable writes being
+                // still present in the journal and not flushed away from it.
+                // So we must sync specific objects here.
+                //
+                // Even more, we have to process "stabilize" request in parts. That is,
+                // we must stabilize all objects which are already synced. Otherwise
+                // they may block objects which are NOT synced yet.
+                return STAB_SPLIT_SYNC;
+            }
+            else if (IS_STABLE(dirty_it->second.state))
+            {
+                break;
+            }
+            // Check previous versions too
+            if (dirty_it == dirty_db.begin())
+            {
+                break;
+            }
+            dirty_it--;
+            if (dirty_it->first.oid != ov.oid)
+            {
+                break;
+            }
         }
+        return STAB_SPLIT_TODO;
     });
     if (r != 1)
     {
@@ -402,7 +416,7 @@ resume_4:
     {
         // Mark all dirty_db entries up to op->version as stable
 #ifdef BLOCKSTORE_DEBUG
-        printf("Stabilize %lx:%lx v%lu\n", v->oid.inode, v->oid.stripe, v->version);
+        printf("Stabilize %jx:%jx v%ju\n", v->oid.inode, v->oid.stripe, v->version);
 #endif
         mark_stable(*v);
     }
@@ -493,7 +507,7 @@ void blockstore_impl_t::mark_stable(obj_ver_id v, bool forget_dirty)
             {
                 // mark_stable should never be called for in-flight or submitted writes
                 printf(
-                    "BUG: Attempt to mark_stable object %lx:%lx v%lu state of which is %x\n",
+                    "BUG: Attempt to mark_stable object %jx:%jx v%ju state of which is %x\n",
                     dirty_it->first.oid.inode, dirty_it->first.oid.stripe, dirty_it->first.version,
                     dirty_it->second.state
                 );
@@ -537,6 +551,5 @@ void blockstore_impl_t::mark_stable(obj_ver_id v, bool forget_dirty)
         unstab_it->second <= v.version)
     {
         unstable_writes.erase(unstab_it);
-        unstable_count_changed = true;
     }
 }

src/blockstore_sync.cpp

@@ -92,8 +92,7 @@ int blockstore_impl_t::continue_sync(blockstore_op_t *op)
             }
         }
         else if (!space_check.check_available(op, PRIV(op)->sync_big_writes.size(),
-            sizeof(journal_entry_big_write) + dsk.clean_entry_bitmap_size,
+            sizeof(journal_entry_big_write) + dsk.clean_entry_bitmap_size, 0))
-            (unstable_writes.size()+unstable_unsynced)*journal.block_size))
         {
             return 0;
         }
@@ -116,11 +115,14 @@ int blockstore_impl_t::continue_sync(blockstore_op_t *op)
                 journal, (dirty_entry.state & BS_ST_INSTANT) ? JE_BIG_WRITE_INSTANT : JE_BIG_WRITE,
                 sizeof(journal_entry_big_write) + dyn_size
             );
-            dirty_entry.journal_sector = journal.sector_info[journal.cur_sector].offset;
+            auto jsec = dirty_entry.journal_sector = journal.sector_info[journal.cur_sector].offset;
+            assert(journal.next_free >= journal.used_start
+                ? (jsec >= journal.used_start && jsec < journal.next_free)
+                : (jsec >= journal.used_start || jsec < journal.next_free));
             journal.used_sectors[journal.sector_info[journal.cur_sector].offset]++;
 #ifdef BLOCKSTORE_DEBUG
             printf(
-                "journal offset %08lx is used by %lx:%lx v%lu (%lu refs)\n",
+                "journal offset %08jx is used by %jx:%jx v%ju (%ju refs)\n",
                 dirty_entry.journal_sector, it->oid.inode, it->oid.stripe, it->version,
                 journal.used_sectors[journal.sector_info[journal.cur_sector].offset]
             );
@@ -174,7 +176,7 @@ void blockstore_impl_t::ack_sync(blockstore_op_t *op)
     for (auto it = PRIV(op)->sync_big_writes.begin(); it != PRIV(op)->sync_big_writes.end(); it++)
     {
 #ifdef BLOCKSTORE_DEBUG
-        printf("Ack sync big %lx:%lx v%lu\n", it->oid.inode, it->oid.stripe, it->version);
+        printf("Ack sync big %jx:%jx v%ju\n", it->oid.inode, it->oid.stripe, it->version);
 #endif
         auto & unstab = unstable_writes[it->oid];
         unstab = unstab < it->version ? it->version : unstab;
@@ -202,7 +204,7 @@ void blockstore_impl_t::ack_sync(blockstore_op_t *op)
     for (auto it = PRIV(op)->sync_small_writes.begin(); it != PRIV(op)->sync_small_writes.end(); it++)
     {
 #ifdef BLOCKSTORE_DEBUG
-        printf("Ack sync small %lx:%lx v%lu\n", it->oid.inode, it->oid.stripe, it->version);
+        printf("Ack sync small %jx:%jx v%ju\n", it->oid.inode, it->oid.stripe, it->version);
 #endif
         auto & unstab = unstable_writes[it->oid];
         unstab = unstab < it->version ? it->version : unstab;

src/blockstore_write.cpp

@@ -85,7 +85,7 @@ bool blockstore_impl_t::enqueue_write(blockstore_op_t *op)
             // It's allowed to write versions with low numbers over deletes
             // However, we have to flush those deletes first as we use version number for ordering
 #ifdef BLOCKSTORE_DEBUG
-            printf("Write %lx:%lx v%lu over delete (real v%lu) offset=%u len=%u\n", op->oid.inode, op->oid.stripe, version, op->version, op->offset, op->len);
+            printf("Write %jx:%jx v%ju over delete (real v%ju) offset=%u len=%u\n", op->oid.inode, op->oid.stripe, version, op->version, op->offset, op->len);
 #endif
             wait_del = true;
             PRIV(op)->real_version = op->version;
@@ -95,11 +95,13 @@ bool blockstore_impl_t::enqueue_write(blockstore_op_t *op)
                 // Issue an additional sync so the delete reaches the journal
                 blockstore_op_t *sync_op = new blockstore_op_t;
                 sync_op->opcode = BS_OP_SYNC;
-                sync_op->callback = [this, op](blockstore_op_t *sync_op)
+                sync_op->oid = op->oid;
+                sync_op->version = op->version;
+                sync_op->callback = [this](blockstore_op_t *sync_op)
                 {
                     flusher->unshift_flush((obj_ver_id){
-                        .oid = op->oid,
+                        .oid = sync_op->oid,
-                        .version = op->version-1,
+                        .version = sync_op->version-1,
                     }, true);
                     delete sync_op;
                 };
@@ -117,7 +119,7 @@ bool blockstore_impl_t::enqueue_write(blockstore_op_t *op)
         {
             // Invalid version requested
 #ifdef BLOCKSTORE_DEBUG
-            printf("Write %lx:%lx v%lu requested, but we already have v%lu\n", op->oid.inode, op->oid.stripe, op->version, version);
+            printf("Write %jx:%jx v%ju requested, but we already have v%ju\n", op->oid.inode, op->oid.stripe, op->version, version);
 #endif
             op->retval = -EEXIST;
             if (!is_del && alloc_dyn_data)
@@ -144,9 +146,9 @@ bool blockstore_impl_t::enqueue_write(blockstore_op_t *op)
         unsynced_queued_ops++;
 #ifdef BLOCKSTORE_DEBUG
     if (is_del)
-        printf("Delete %lx:%lx v%lu\n", op->oid.inode, op->oid.stripe, op->version);
+        printf("Delete %jx:%jx v%ju\n", op->oid.inode, op->oid.stripe, op->version);
     else if (!wait_del)
-        printf("Write %lx:%lx v%lu offset=%u len=%u\n", op->oid.inode, op->oid.stripe, op->version, op->offset, op->len);
+        printf("Write %jx:%jx v%ju offset=%u len=%u\n", op->oid.inode, op->oid.stripe, op->version, op->offset, op->len);
 #endif
     // No strict need to add it into dirty_db here except maybe for listings to return
     // correct data when there are inflight operations in the queue
@@ -286,7 +288,7 @@ int blockstore_impl_t::dequeue_write(blockstore_op_t *op)
         }
         // Restore original low version number for unblocked operations
 #ifdef BLOCKSTORE_DEBUG
-        printf("Restoring %lx:%lx version: v%lu -> v%lu\n", op->oid.inode, op->oid.stripe, op->version, PRIV(op)->real_version);
+        printf("Restoring %jx:%jx version: v%ju -> v%ju\n", op->oid.inode, op->oid.stripe, op->version, PRIV(op)->real_version);
 #endif
         auto prev_it = dirty_it;
         if (prev_it != dirty_db.begin())
@@ -296,7 +298,7 @@ int blockstore_impl_t::dequeue_write(blockstore_op_t *op)
             {
                 // Original version is still invalid
                 // All subsequent writes to the same object must be canceled too
-                printf("Tried to write %lx:%lx v%lu after delete (old version v%lu), but already have v%lu\n",
+                printf("Tried to write %jx:%jx v%ju after delete (old version v%ju), but already have v%ju\n",
                     op->oid.inode, op->oid.stripe, PRIV(op)->real_version, op->version, prev_it->first.version);
                 cancel_all_writes(op, dirty_it, -EEXIST);
                 return 2;
@@ -320,7 +322,7 @@ int blockstore_impl_t::dequeue_write(blockstore_op_t *op)
         blockstore_journal_check_t space_check(this);
         if (!space_check.check_available(op, unsynced_big_write_count + 1,
             sizeof(journal_entry_big_write) + dsk.clean_dyn_size,
-            (unstable_writes.size()+unstable_unsynced)*journal.block_size))
+            (unstable_writes.size()+unstable_unsynced+((dirty_it->second.state & BS_ST_INSTANT) ? 0 : 1))*journal.block_size))
         {
             return 0;
         }
@@ -348,8 +350,8 @@ int blockstore_impl_t::dequeue_write(blockstore_op_t *op)
             if (entry->oid.inode || entry->oid.stripe || entry->version)
             {
                 printf(
-                    "Fatal error (metadata corruption or bug): tried to write object %lx:%lx v%lu"
+                    "Fatal error (metadata corruption or bug): tried to write object %jx:%jx v%ju"
-                    " over a non-zero metadata entry %lu with %lx:%lx v%lu\n", op->oid.inode,
+                    " over a non-zero metadata entry %ju with %jx:%jx v%ju\n", op->oid.inode,
                     op->oid.stripe, op->version, loc, entry->oid.inode, entry->oid.stripe, entry->version
                 );
                 exit(1);
@@ -361,7 +363,7 @@ int blockstore_impl_t::dequeue_write(blockstore_op_t *op)
         dirty_it->second.state = (dirty_it->second.state & ~BS_ST_WORKFLOW_MASK) | BS_ST_SUBMITTED;
 #ifdef BLOCKSTORE_DEBUG
         printf(
-            "Allocate block %lu for %lx:%lx v%lu\n",
+            "Allocate block %ju for %jx:%jx v%ju\n",
             loc, op->oid.inode, op->oid.stripe, op->version
         );
 #endif
@@ -372,13 +374,13 @@ int blockstore_impl_t::dequeue_write(blockstore_op_t *op)
         int vcnt = 0;
         if (stripe_offset)
         {
-            PRIV(op)->iov_zerofill[vcnt++] = (struct iovec){ zero_object, stripe_offset };
+            PRIV(op)->iov_zerofill[vcnt++] = (struct iovec){ zero_object, (size_t)stripe_offset };
         }
         PRIV(op)->iov_zerofill[vcnt++] = (struct iovec){ op->buf, op->len };
         if (stripe_end)
         {
             stripe_end = dsk.bitmap_granularity - stripe_end;
-            PRIV(op)->iov_zerofill[vcnt++] = (struct iovec){ zero_object, stripe_end };
+            PRIV(op)->iov_zerofill[vcnt++] = (struct iovec){ zero_object, (size_t)stripe_end };
         }
         data->iov.iov_len = op->len + stripe_offset + stripe_end; // to check it in the callback
         data->callback = [this, op](ring_data_t *data) { handle_write_event(data, op); };
@@ -412,7 +414,7 @@ int blockstore_impl_t::dequeue_write(blockstore_op_t *op)
                 sizeof(journal_entry_big_write) + dsk.clean_dyn_size, 0)
             || !space_check.check_available(op, 1,
                 sizeof(journal_entry_small_write) + dyn_size,
-                op->len + (unstable_writes.size()+unstable_unsynced)*journal.block_size))
+                op->len + (unstable_writes.size()+unstable_unsynced+((dirty_it->second.state & BS_ST_INSTANT) ? 0 : 1))*journal.block_size))
         {
             return 0;
         }
@@ -436,11 +438,23 @@ int blockstore_impl_t::dequeue_write(blockstore_op_t *op)
             journal, op->opcode == BS_OP_WRITE_STABLE ? JE_SMALL_WRITE_INSTANT : JE_SMALL_WRITE,
             sizeof(journal_entry_small_write) + dyn_size
         );
-        dirty_it->second.journal_sector = journal.sector_info[journal.cur_sector].offset;
+        auto jsec = dirty_it->second.journal_sector = journal.sector_info[journal.cur_sector].offset;
+        if (!(journal.next_free >= journal.used_start
+            ? (jsec >= journal.used_start && jsec < journal.next_free)
+            : (jsec >= journal.used_start || jsec < journal.next_free)))
+        {
+            printf(
+                "BUG: journal offset %08jx is used by %jx:%jx v%ju (%ju refs) BUT used_start=%jx next_free=%jx\n",
+                dirty_it->second.journal_sector, dirty_it->first.oid.inode, dirty_it->first.oid.stripe, dirty_it->first.version,
+                journal.used_sectors[journal.sector_info[journal.cur_sector].offset],
+                journal.used_start, journal.next_free
+            );
+            abort();
+        }
         journal.used_sectors[journal.sector_info[journal.cur_sector].offset]++;
 #ifdef BLOCKSTORE_DEBUG
         printf(
-            "journal offset %08lx is used by %lx:%lx v%lu (%lu refs)\n",
+            "journal offset %08jx is used by %jx:%jx v%ju (%ju refs)\n",
             dirty_it->second.journal_sector, dirty_it->first.oid.inode, dirty_it->first.oid.stripe, dirty_it->first.version,
             journal.used_sectors[journal.sector_info[journal.cur_sector].offset]
         );
@@ -454,8 +468,8 @@ int blockstore_impl_t::dequeue_write(blockstore_op_t *op)
                 journal_used_it->first < next_next_free + op->len)
             {
                 printf(
-                    "BUG: Attempt to overwrite used offset (%lx, %lu refs) of the journal with the object %lx:%lx v%lu: data at %lx, len %x!"
+                    "BUG: Attempt to overwrite used offset (%jx, %ju refs) of the journal with the object %jx:%jx v%ju: data at %jx, len %x!"
-                    " Journal used_start=%08lx (%lu refs), next_free=%08lx, dirty_start=%08lx\n",
+                    " Journal used_start=%08jx (%ju refs), next_free=%08jx, dirty_start=%08jx\n",
                     journal_used_it->first, journal_used_it->second, op->oid.inode, op->oid.stripe, op->version, next_next_free, op->len,
                     journal.used_start, journal.used_sectors[journal.used_start], journal.next_free, journal.dirty_start
                 );
@@ -463,7 +477,7 @@ int blockstore_impl_t::dequeue_write(blockstore_op_t *op)
             }
         }
         // double check that next_free doesn't cross used_start from the left
-        assert(journal.next_free >= journal.used_start || next_next_free < journal.used_start);
+        assert(journal.next_free >= journal.used_start && next_next_free >= journal.next_free || next_next_free < journal.used_start);
         journal.next_free = next_next_free;
         je->oid = op->oid;
         je->version = op->version;
@@ -505,7 +519,7 @@ int blockstore_impl_t::dequeue_write(blockstore_op_t *op)
         if (next_next_free >= journal.len)
             next_next_free = dsk.journal_block_size;
         // double check that next_free doesn't cross used_start from the left
-        assert(journal.next_free >= journal.used_start || next_next_free < journal.used_start);
+        assert(journal.next_free >= journal.used_start && next_next_free >= journal.next_free || next_next_free < journal.used_start);
         journal.next_free = next_next_free;
         if (!(dirty_it->second.state & BS_ST_INSTANT))
         {
@@ -549,7 +563,7 @@ resume_2:
         uint64_t dyn_size = dsk.dirty_dyn_size(op->offset, op->len);
         blockstore_journal_check_t space_check(this);
         if (!space_check.check_available(op, 1, sizeof(journal_entry_big_write) + dyn_size,
-            (unstable_writes.size()+unstable_unsynced)*journal.block_size))
+            (unstable_writes.size()+unstable_unsynced+((dirty_it->second.state & BS_ST_INSTANT) ? 0 : 1))*journal.block_size))
         {
             return 0;
         }
@@ -558,11 +572,23 @@ resume_2:
             journal, op->opcode == BS_OP_WRITE_STABLE ? JE_BIG_WRITE_INSTANT : JE_BIG_WRITE,
             sizeof(journal_entry_big_write) + dyn_size
         );
-        dirty_it->second.journal_sector = journal.sector_info[journal.cur_sector].offset;
+        auto jsec = dirty_it->second.journal_sector = journal.sector_info[journal.cur_sector].offset;
+        if (!(journal.next_free >= journal.used_start
+            ? (jsec >= journal.used_start && jsec < journal.next_free)
+            : (jsec >= journal.used_start || jsec < journal.next_free)))
+        {
+            printf(
+                "BUG: journal offset %08jx is used by %jx:%jx v%ju (%ju refs) BUT used_start=%jx next_free=%jx\n",
+                dirty_it->second.journal_sector, dirty_it->first.oid.inode, dirty_it->first.oid.stripe, dirty_it->first.version,
+                journal.used_sectors[journal.sector_info[journal.cur_sector].offset],
+                journal.used_start, journal.next_free
+            );
+            abort();
+        }
         journal.used_sectors[journal.sector_info[journal.cur_sector].offset]++;
 #ifdef BLOCKSTORE_DEBUG
         printf(
-            "journal offset %08lx is used by %lx:%lx v%lu (%lu refs)\n",
+            "journal offset %08jx is used by %jx:%jx v%ju (%ju refs)\n",
             journal.sector_info[journal.cur_sector].offset, op->oid.inode, op->oid.stripe, op->version,
             journal.used_sectors[journal.sector_info[journal.cur_sector].offset]
         );
@@ -589,7 +615,7 @@ resume_4:
         });
         assert(dirty_it != dirty_db.end());
 #ifdef BLOCKSTORE_DEBUG
-        printf("Ack write %lx:%lx v%lu = state 0x%x\n", op->oid.inode, op->oid.stripe, op->version, dirty_it->second.state);
+        printf("Ack write %jx:%jx v%ju = state 0x%x\n", op->oid.inode, op->oid.stripe, op->version, dirty_it->second.state);
 #endif
         bool is_big = (dirty_it->second.state & BS_ST_TYPE_MASK) == BS_ST_BIG_WRITE;
         bool imm = is_big ? (immediate_commit == IMMEDIATE_ALL) : (immediate_commit != IMMEDIATE_NONE);
@@ -782,7 +808,7 @@ int blockstore_impl_t::dequeue_del(blockstore_op_t *op)
     journal.used_sectors[journal.sector_info[journal.cur_sector].offset]++;
 #ifdef BLOCKSTORE_DEBUG
     printf(
-        "journal offset %08lx is used by %lx:%lx v%lu (%lu refs)\n",
+        "journal offset %08jx is used by %jx:%jx v%ju (%ju refs)\n",
         dirty_it->second.journal_sector, dirty_it->first.oid.inode, dirty_it->first.oid.stripe, dirty_it->first.version,
         journal.used_sectors[journal.sector_info[journal.cur_sector].offset]
     );

src/cli.cpp

@@ -46,18 +46,21 @@ static const char* help_text =
     "vitastor-cli snap-create [-p|--pool <id|name>] <image>@<snapshot>\n"
     "  Create a snapshot of image <name>. May be used live if only a single writer is active.\n"
     "\n"
-    "vitastor-cli modify <name> [--rename <new-name>] [--resize <size>] [--readonly | --readwrite] [-f|--force]\n"
+    "vitastor-cli modify <name> [--rename <new-name>] [--resize <size>] [--readonly | --readwrite] [-f|--force] [--down-ok]\n"
     "  Rename, resize image or change its readonly status. Images with children can't be made read-write.\n"
     "  If the new size is smaller than the old size, extra data will be purged.\n"
     "  You should resize file system in the image, if present, before shrinking it.\n"
     "  -f|--force  Proceed with shrinking or setting readwrite flag even if the image has children.\n"
+    "  --down-ok   Proceed with shrinking even if some data will be left on unavailable OSDs.\n"
     "\n"
-    "vitastor-cli rm <from> [<to>] [--writers-stopped]\n"
+    "vitastor-cli rm <from> [<to>] [--writers-stopped] [--down-ok]\n"
     "  Remove <from> or all layers between <from> and <to> (<to> must be a child of <from>),\n"
     "  rebasing all their children accordingly. --writers-stopped allows merging to be a bit\n"
     "  more effective in case of a single 'slim' read-write child and 'fat' removed parent:\n"
     "  the child is merged into parent and parent is renamed to child in that case.\n"
     "  In other cases parent layers are always merged into children.\n"
+    "  Other options:\n"
+    "  --down-ok  Continue deletion/merging even if some data will be left on unavailable OSDs.\n"
     "\n"
     "vitastor-cli flatten <layer>\n"
     "  Flatten a layer, i.e. merge data and detach it from parents.\n"
@@ -122,7 +125,7 @@ static const char* help_text =
     "  --parallel_osds M   Work with M osds in parallel when possible (default 4)\n"
     "  --progress 1|0      Report progress (default 1)\n"
     "  --cas 1|0           Use CAS writes for flatten, merge, rm (default is decide automatically)\n"
-    "  --no-color          Disable colored output\n"
+    "  --color 1|0         Enable/disable colored output and CR symbols (default 1 if stdout is a terminal)\n"
     "  --json              JSON output\n"
 ;
 
@@ -171,6 +174,7 @@ static json11::Json::object parse_args(int narg, const char *args[])
                 !strcmp(opt, "readonly") || !strcmp(opt, "readwrite") ||
                 !strcmp(opt, "force") || !strcmp(opt, "reverse") ||
                 !strcmp(opt, "allow-data-loss") || !strcmp(opt, "allow_data_loss") ||
+                !strcmp(opt, "down-ok") || !strcmp(opt, "down_ok") ||
                 !strcmp(opt, "dry-run") || !strcmp(opt, "dry_run") ||
                 !strcmp(opt, "help") || !strcmp(opt, "all") ||
                 (!strcmp(opt, "writers-stopped") || !strcmp(opt, "writers_stopped")) && strcmp("1", args[i+1]) != 0

src/cli_alloc_osd.cpp

@@ -77,7 +77,7 @@ struct alloc_osd_t
                     std::string key = base64_decode(kv["key"].string_value());
                     osd_num_t cur_osd;
                     char null_byte = 0;
-                    int scanned = sscanf(key.c_str() + parent->cli->st_cli.etcd_prefix.length(), "/osd/stats/%lu%c", &cur_osd, &null_byte);
+                    int scanned = sscanf(key.c_str() + parent->cli->st_cli.etcd_prefix.length(), "/osd/stats/%ju%c", &cur_osd, &null_byte);
                     if (scanned != 1 || !cur_osd)
                     {
                         fprintf(stderr, "Invalid key in etcd: %s\n", key.c_str());

src/cli_common.cpp

@@ -1,6 +1,7 @@
 // Copyright (c) Vitaliy Filippov, 2019+
 // License: VNPL-1.1 (see README.md for details)
 
+#include <unistd.h>
 #include "str_util.h"
 #include "cluster_client.h"
 #include "cli.h"
@@ -11,7 +12,7 @@ void cli_tool_t::change_parent(inode_t cur, inode_t new_parent, cli_result_t *re
     if (cur_cfg_it == cli->st_cli.inode_config.end())
     {
         char buf[128];
-        snprintf(buf, 128, "Inode 0x%lx disappeared", cur);
+        snprintf(buf, 128, "Inode 0x%jx disappeared", cur);
         *result = (cli_result_t){ .err = EIO, .text = buf };
         return;
     }
@@ -113,7 +114,12 @@ void cli_tool_t::parse_config(json11::Json::object & cfg)
         else
             kv_it++;
     }
-    color = !cfg["no_color"].bool_value();
+    if (cfg.find("no_color") != cfg.end())
+        color = !cfg["no_color"].bool_value();
+    else if (cfg.find("color") != cfg.end())
+        color = cfg["color"].bool_value();
+    else
+        color = isatty(1);
     json_output = cfg["json"].bool_value();
     iodepth = cfg["iodepth"].uint64_value();
     if (!iodepth)

src/cli_describe.cpp

@@ -160,14 +160,14 @@ struct cli_describe_t
                 if (op->reply.hdr.retval < 0)
                 {
                     fprintf(
-                        stderr, "Failed to describe objects on OSD %lu (retval=%ld)\n",
+                        stderr, "Failed to describe objects on OSD %ju (retval=%jd)\n",
                         osd_num, op->reply.hdr.retval
                     );
                 }
                 else if (op->reply.describe.result_bytes != op->reply.hdr.retval * sizeof(osd_reply_describe_item_t))
                 {
                     fprintf(
-                        stderr, "Invalid response size from OSD %lu (expected %lu bytes, got %lu bytes)\n",
+                        stderr, "Invalid response size from OSD %ju (expected %ju bytes, got %ju bytes)\n",
                         osd_num, op->reply.hdr.retval * sizeof(osd_reply_describe_item_t), op->reply.describe.result_bytes
                     );
                 }
@@ -178,11 +178,11 @@ struct cli_describe_t
                     {
                         if (!parent->json_output || parent->is_command_line)
                         {
-#define FMT "{\"inode\":\"0x%lx\",\"stripe\":\"0x%lx\",\"part\":%u,\"osd_num\":%lu%s%s%s}"
+#define FMT "{\"inode\":\"0x%jx\",\"stripe\":\"0x%jx\",\"part\":%u,\"osd_num\":%ju%s%s%s}"
                             printf(
                                 (parent->json_output
                                     ? (count > 0 ? ",\n  " FMT : "  " FMT)
-                                    : "%lx:%lx part %u on OSD %lu%s%s%s\n"),
+                                    : "%jx:%jx part %u on OSD %ju%s%s%s\n"),
 #undef FMT
                                 items[i].inode, items[i].stripe,
                                 items[i].role, items[i].osd_num,

src/cli_df.cpp

@@ -82,7 +82,7 @@ resume_1:
             // osd ID
             osd_num_t osd_num;
             char null_byte = 0;
-            int scanned = sscanf(kv.key.substr(parent->cli->st_cli.etcd_prefix.length()).c_str(), "/osd/stats/%lu%c", &osd_num, &null_byte);
+            int scanned = sscanf(kv.key.substr(parent->cli->st_cli.etcd_prefix.length()).c_str(), "/osd/stats/%ju%c", &osd_num, &null_byte);
             if (scanned != 1 || !osd_num || osd_num >= POOL_ID_MAX)
             {
                 fprintf(stderr, "Invalid key in etcd: %s\n", kv.key.c_str());

src/cli_fix.cpp

@@ -136,7 +136,7 @@ struct cli_fix_t
             auto pool_cfg_it = parent->cli->st_cli.pool_config.find(INODE_POOL(obj.inode));
             if (pool_cfg_it == parent->cli->st_cli.pool_config.end())
             {
-                fprintf(stderr, "Object %lx:%lx is from unknown pool\n", obj.inode, obj.stripe);
+                fprintf(stderr, "Object %jx:%jx is from unknown pool\n", obj.inode, obj.stripe);
                 continue;
             }
             auto & pool_cfg = pool_cfg_it->second;
@@ -146,7 +146,7 @@ struct cli_fix_t
                 !pg_it->second.cur_primary || !(pg_it->second.cur_state & PG_ACTIVE))
             {
                 fprintf(
-                    stderr, "Object %lx:%lx is from PG %u/%u which is not currently active\n",
+                    stderr, "Object %jx:%jx is from PG %u/%u which is not currently active\n",
                     obj.inode, obj.stripe, pool_cfg_it->first, pg_num
                 );
                 continue;
@@ -171,7 +171,7 @@ struct cli_fix_t
             {
                 if (op->reply.hdr.retval < 0 || op->reply.describe.result_bytes != op->reply.hdr.retval * sizeof(osd_reply_describe_item_t))
                 {
-                    fprintf(stderr, "Failed to describe objects on OSD %lu (retval=%ld)\n", primary_osd, op->reply.hdr.retval);
+                    fprintf(stderr, "Failed to describe objects on OSD %ju (retval=%jd)\n", primary_osd, op->reply.hdr.retval);
                     parent->waiting--;
                     loop();
                 }
@@ -209,7 +209,7 @@ struct cli_fix_t
                                 if (rm_op->reply.hdr.retval < 0)
                                 {
                                     fprintf(
-                                        stderr, "Failed to remove object %lx:%lx from OSD %lu (retval=%ld)\n",
+                                        stderr, "Failed to remove object %jx:%jx from OSD %ju (retval=%jd)\n",
                                         rm_op->req.sec_del.oid.inode, rm_op->req.sec_del.oid.stripe,
                                         rm_osd_num, rm_op->reply.hdr.retval
                                     );
@@ -226,7 +226,7 @@ struct cli_fix_t
                                 else
                                 {
                                     printf(
-                                        "Removed %lx:%lx (part %lu) from OSD %lu\n",
+                                        "Removed %jx:%jx (part %ju) from OSD %ju\n",
                                         rm_op->req.sec_del.oid.inode, rm_op->req.sec_del.oid.stripe & ~STRIPE_MASK,
                                         rm_op->req.sec_del.oid.stripe & STRIPE_MASK, rm_osd_num
                                     );
@@ -254,7 +254,7 @@ struct cli_fix_t
                                         if (scrub_op->reply.hdr.retval < 0 && scrub_op->reply.hdr.retval != -ENOENT)
                                         {
                                             fprintf(
-                                                stderr, "Failed to scrub %lx:%lx on OSD %lu (retval=%ld)\n",
+                                                stderr, "Failed to scrub %jx:%jx on OSD %ju (retval=%jd)\n",
                                                 obj.inode, obj.stripe, primary_osd, scrub_op->reply.hdr.retval
                                             );
                                         }

src/cli_ls.cpp

@@ -150,7 +150,7 @@ resume_1:
             inode_t only_inode_num;
             char null_byte = 0;
             int scanned = sscanf(kv.key.substr(parent->cli->st_cli.etcd_prefix.length()).c_str(),
-                "/inode/stats/%u/%lu%c", &pool_id, &only_inode_num, &null_byte);
+                "/inode/stats/%u/%ju%c", &pool_id, &only_inode_num, &null_byte);
             if (scanned != 2 || !pool_id || pool_id >= POOL_ID_MAX || INODE_POOL(only_inode_num) != 0)
             {
                 fprintf(stderr, "Invalid key in etcd: %s\n", kv.key.c_str());
@@ -456,7 +456,7 @@ std::string format_lat(uint64_t lat)
     char buf[256];
     int l = 0;
     if (lat < 100)
-        l = snprintf(buf, sizeof(buf), "%lu us", lat);
+        l = snprintf(buf, sizeof(buf), "%ju us", lat);
     else if (lat < 500000)
         l = snprintf(buf, sizeof(buf), "%.2f ms", (double)lat/1000);
     else

src/cli_merge.cpp

@@ -202,7 +202,7 @@ struct snap_merger_t
         if (parent->progress)
         {
             printf(
-                "Merging %ld layer(s) into target %s%s (inode %lu in pool %u)\n",
+                "Merging %zd layer(s) into target %s%s (inode %ju in pool %u)\n",
                 sources.size(), target_cfg->name.c_str(),
                 use_cas ? " online (with CAS)" : "", INODE_NO_POOL(target), INODE_POOL(target)
             );
@@ -275,7 +275,9 @@ struct snap_merger_t
                 processed++;
                 if (parent->progress && !(processed % 128))
                 {
-                    printf("\rFiltering target blocks: %lu/%lu", processed, to_process);
+                    fprintf(stderr, parent->color
+                        ? "\rFiltering target blocks: %ju/%ju"
+                        : "Filtering target blocks: %ju/%ju\n", processed, to_process);
                 }
             }
             if (in_flight > 0 || oit != merge_offsets.end())
@@ -285,7 +287,9 @@ struct snap_merger_t
             }
             if (parent->progress)
             {
-                printf("\r%lu full blocks of target filtered out\n", to_process-merge_offsets.size());
+                fprintf(stderr, parent->color
+                    ? "\r%ju full blocks of target filtered out\n"
+                    : "%ju full blocks of target filtered out\n", to_process-merge_offsets.size());
             }
         }
         state = 3;
@@ -320,7 +324,9 @@ struct snap_merger_t
             processed++;
             if (parent->progress && !(processed % 128))
             {
-                printf("\rOverwriting blocks: %lu/%lu", processed, to_process);
+                fprintf(stderr, parent->color
+                    ? "\rOverwriting blocks: %ju/%ju"
+                    : "Overwriting blocks: %ju/%ju\n", processed, to_process);
             }
         }
         if (in_flight == 0 && rwo_error.size())
@@ -339,7 +345,9 @@ struct snap_merger_t
         }
         if (parent->progress)
         {
-            printf("\rOverwriting blocks: %lu/%lu\n", to_process, to_process);
+            fprintf(stderr, parent->color
+                ? "\rOverwriting blocks: %ju/%ju\n"
+                : "Overwriting blocks: %ju/%ju\n", to_process, to_process);
         }
         // Done
         result = (cli_result_t){ .text = "Done, layers from "+from_name+" to "+to_name+" merged into "+target_name };
@@ -384,7 +392,7 @@ struct snap_merger_t
                         auto & name = parent->cli->st_cli.inode_config.at(src).name;
                         if (parent->progress)
                         {
-                            printf("Got listing of layer %s (inode %lu in pool %u)\n", name.c_str(), INODE_NO_POOL(src), INODE_POOL(src));
+                            printf("Got listing of layer %s (inode %ju in pool %u)\n", name.c_str(), INODE_NO_POOL(src), INODE_POOL(src));
                         }
                         if (delete_source)
                         {
@@ -416,7 +424,7 @@ struct snap_merger_t
         {
             if (op->retval < 0)
             {
-                fprintf(stderr, "error reading target bitmap at offset %lx: %s\n", op->offset, strerror(-op->retval));
+                fprintf(stderr, "error reading target bitmap at offset %jx: %s\n", op->offset, strerror(-op->retval));
             }
             else
             {
@@ -571,7 +579,7 @@ struct snap_merger_t
         {
             if (subop->retval != 0)
             {
-                fprintf(stderr, "error deleting from layer 0x%lx at offset %lx: %s", subop->inode, subop->offset, strerror(-subop->retval));
+                fprintf(stderr, "error deleting from layer 0x%jx at offset %jx: %s", subop->inode, subop->offset, strerror(-subop->retval));
             }
             delete subop;
         };
@@ -620,7 +628,7 @@ struct snap_merger_t
             if (rwo->error_code)
             {
                 char buf[1024];
-                snprintf(buf, 1024, "Error %s target at offset %lx: %s",
+                snprintf(buf, 1024, "Error %s target at offset %jx: %s",
                     rwo->error_read ? "reading" : "writing", rwo->error_offset, strerror(rwo->error_code));
                 rwo_error = std::string(buf);
             }

src/cli_modify.cpp

@@ -15,6 +15,7 @@ struct image_changer_t
     uint64_t new_size = 0;
     bool force_size = false, inc_size = false;
     bool set_readonly = false, set_readwrite = false, force = false;
+    bool down_ok = false;
     // interval between fsyncs
     int fsync_interval = 128;
 
@@ -105,6 +106,7 @@ struct image_changer_t
                     { "pool", (uint64_t)INODE_POOL(inode_num) },
                     { "fsync-interval", fsync_interval },
                     { "min-offset", ((new_size+4095)/4096)*4096 },
+                    { "down-ok", down_ok },
                 });
 resume_1:
                 while (!cb(result))
@@ -240,6 +242,7 @@ std::function<bool(cli_result_t &)> cli_tool_t::start_modify(json11::Json cfg)
     changer->fsync_interval = cfg["fsync_interval"].uint64_value();
     if (!changer->fsync_interval)
         changer->fsync_interval = 128;
+    changer->down_ok = cfg["down_ok"].bool_value();
     // FIXME Check that the image doesn't have children when shrinking
     return [changer](cli_result_t & result)
     {

src/cli_rm.cpp

@@ -53,6 +53,8 @@ struct snap_remover_t
     int use_cas = 1;
     // interval between fsyncs
     int fsync_interval = 128;
+    // ignore deletion errors
+    bool down_ok = false;
 
     std::map<inode_t,int> sources;
     std::map<inode_t,uint64_t> inode_used;
@@ -291,7 +293,7 @@ resume_100:
             if (it == parent->cli->st_cli.inode_config.end())
             {
                 char buf[1024];
-                snprintf(buf, 1024, "Parent inode of layer %s (id 0x%lx) not found", cur->name.c_str(), cur->parent_id);
+                snprintf(buf, 1024, "Parent inode of layer %s (id 0x%jx) not found", cur->name.c_str(), cur->parent_id);
                 state = 100;
                 return;
             }
@@ -384,7 +386,7 @@ resume_100:
                 pool_id_t pool_id = 0;
                 inode_t inode = 0;
                 char null_byte = 0;
-                int scanned = sscanf(kv.key.c_str() + parent->cli->st_cli.etcd_prefix.length()+13, "%u/%lu%c", &pool_id, &inode, &null_byte);
+                int scanned = sscanf(kv.key.c_str() + parent->cli->st_cli.etcd_prefix.length()+13, "%u/%ju%c", &pool_id, &inode, &null_byte);
                 if (scanned != 2 || !inode)
                 {
                     result = (cli_result_t){ .err = EIO, .text = "Bad key returned from etcd: "+kv.key };
@@ -439,7 +441,7 @@ resume_100:
         if (child_it == parent->cli->st_cli.inode_config.end())
         {
             char buf[1024];
-            snprintf(buf, 1024, "Inode 0x%lx disappeared", inverse_child);
+            snprintf(buf, 1024, "Inode 0x%jx disappeared", inverse_child);
             result = (cli_result_t){ .err = EIO, .text = std::string(buf) };
             state = 100;
             return;
@@ -448,7 +450,7 @@ resume_100:
         if (target_it == parent->cli->st_cli.inode_config.end())
         {
             char buf[1024];
-            snprintf(buf, 1024, "Inode 0x%lx disappeared", inverse_parent);
+            snprintf(buf, 1024, "Inode 0x%jx disappeared", inverse_parent);
             result = (cli_result_t){ .err = EIO, .text = std::string(buf) };
             state = 100;
             return;
@@ -576,7 +578,7 @@ resume_100:
         if (cur_cfg_it == parent->cli->st_cli.inode_config.end())
         {
             char buf[1024];
-            snprintf(buf, 1024, "Inode 0x%lx disappeared", cur);
+            snprintf(buf, 1024, "Inode 0x%jx disappeared", cur);
             result = (cli_result_t){ .err = EIO, .text = std::string(buf) };
             state = 100;
             return;
@@ -640,7 +642,7 @@ resume_100:
         if (child_it == parent->cli->st_cli.inode_config.end())
         {
             char buf[1024];
-            snprintf(buf, 1024, "Inode 0x%lx disappeared", child_inode);
+            snprintf(buf, 1024, "Inode 0x%jx disappeared", child_inode);
             result = (cli_result_t){ .err = EIO, .text = std::string(buf) };
             state = 100;
             return;
@@ -649,7 +651,7 @@ resume_100:
         if (target_it == parent->cli->st_cli.inode_config.end())
         {
             char buf[1024];
-            snprintf(buf, 1024, "Inode 0x%lx disappeared", target_inode);
+            snprintf(buf, 1024, "Inode 0x%jx disappeared", target_inode);
             result = (cli_result_t){ .err = EIO, .text = std::string(buf) };
             state = 100;
             return;
@@ -670,7 +672,7 @@ resume_100:
         if (source == parent->cli->st_cli.inode_config.end())
         {
             char buf[1024];
-            snprintf(buf, 1024, "Inode 0x%lx disappeared", inode);
+            snprintf(buf, 1024, "Inode 0x%jx disappeared", inode);
             result = (cli_result_t){ .err = EIO, .text = std::string(buf) };
             state = 100;
             return;
@@ -679,6 +681,7 @@ resume_100:
             { "inode", inode },
             { "pool", (uint64_t)INODE_POOL(inode) },
             { "fsync-interval", fsync_interval },
+            { "down-ok", down_ok },
         });
     }
 };
@@ -690,6 +693,7 @@ std::function<bool(cli_result_t &)> cli_tool_t::start_rm(json11::Json cfg)
     snap_remover->from_name = cfg["from"].string_value();
     snap_remover->to_name = cfg["to"].string_value();
     snap_remover->fsync_interval = cfg["fsync_interval"].uint64_value();
+    snap_remover->down_ok = cfg["down_ok"].bool_value();
     if (!snap_remover->fsync_interval)
         snap_remover->fsync_interval = 128;
     if (!cfg["cas"].is_null())

src/cli_rm_data.cpp

@@ -25,6 +25,7 @@ struct rm_inode_t
     uint64_t inode = 0;
     pool_id_t pool_id = 0;
     uint64_t min_offset = 0;
+    bool down_ok = false;
 
     cli_tool_t *parent = NULL;
     inode_list_t *lister = NULL;
@@ -95,7 +96,7 @@ struct rm_inode_t
             fprintf(stderr, "Some data may remain after delete on OSDs which are currently down: ");
             for (int i = 0; i < inactive_osds.size(); i++)
             {
-                fprintf(stderr, i > 0 ? ", %lu" : "%lu", inactive_osds[i]);
+                fprintf(stderr, i > 0 ? ", %ju" : "%ju", inactive_osds[i]);
             }
             fprintf(stderr, "\n");
         }
@@ -138,7 +139,7 @@ struct rm_inode_t
                     cur_list->in_flight--;
                     if (op->reply.hdr.retval < 0)
                     {
-                        fprintf(stderr, "Failed to remove object %lx:%lx from PG %u (OSD %lu) (retval=%ld)\n",
+                        fprintf(stderr, "Failed to remove object %jx:%jx from PG %u (OSD %ju) (retval=%jd)\n",
                             op->req.rw.inode, op->req.rw.offset,
                             cur_list->pg_num, cur_list->rm_osd_num, op->reply.hdr.retval);
                         error_count++;
@@ -174,7 +175,7 @@ struct rm_inode_t
                 cur_list->synced = true;
                 if (op->reply.hdr.retval < 0)
                 {
-                    fprintf(stderr, "Failed to sync OSD %lu (retval=%ld)\n",
+                    fprintf(stderr, "Failed to sync OSD %ju (retval=%jd)\n",
                         cur_list->rm_osd_num, op->reply.hdr.retval);
                     error_count++;
                 }
@@ -212,7 +213,9 @@ struct rm_inode_t
         }
         if (parent->progress && total_count > 0 && total_done*1000/total_count != total_prev_pct)
         {
-            fprintf(stderr, "\rRemoved %lu/%lu objects, %lu more PGs to list...", total_done, total_count, pgs_to_list);
+            fprintf(stderr, parent->color
+                ? "\rRemoved %ju/%ju objects, %ju more PGs to list..."
+                : "Removed %ju/%ju objects, %ju more PGs to list...\n", total_done, total_count, pgs_to_list);
             total_prev_pct = total_done*1000/total_count;
         }
         if (lists_done && !lists.size())
@@ -221,17 +224,18 @@ struct rm_inode_t
             {
                 fprintf(stderr, "\n");
             }
-            if (parent->progress && (total_done < total_count || inactive_osds.size() > 0 || error_count > 0))
+            bool is_error = (total_done < total_count || inactive_osds.size() > 0 || error_count > 0);
+            if (parent->progress && is_error)
             {
                 fprintf(
-                    stderr, "Warning: Pool:%u,ID:%lu inode data may not have been fully removed.\n"
+                    stderr, "Warning: Pool:%u,ID:%ju inode data may not have been fully removed.\n"
-                    " Use `vitastor-cli rm-data --pool %u --inode %lu` if you encounter it in listings.\n",
+                    "Use `vitastor-cli rm-data --pool %u --inode %ju` if you encounter it in listings.\n",
                     pool_id, INODE_NO_POOL(inode), pool_id, INODE_NO_POOL(inode)
                 );
             }
             result = (cli_result_t){
-                .err = error_count > 0 ? EIO : 0,
+                .err = is_error && !down_ok ? EIO : 0,
-                .text = error_count > 0 ? "Some blocks were not removed" : (
+                .text = is_error ? "Some blocks were not removed" : (
                     "Done, inode "+std::to_string(INODE_NO_POOL(inode))+" from pool "+
                     std::to_string(pool_id)+" removed"),
                 .data = json11::Json::object {
@@ -280,6 +284,7 @@ std::function<bool(cli_result_t &)> cli_tool_t::start_rm_data(json11::Json cfg)
     {
         remover->inode = (remover->inode & (((uint64_t)1 << (64-POOL_ID_BITS)) - 1)) | (((uint64_t)remover->pool_id) << (64-POOL_ID_BITS));
     }
+    remover->down_ok = cfg["down_ok"].bool_value();
     remover->pool_id = INODE_POOL(remover->inode);
     remover->min_offset = cfg["min_offset"].uint64_value();
     return [remover](cli_result_t & result)

src/cli_status.cpp

@@ -106,7 +106,7 @@ resume_2:
             if (etcd_states[i]["error"].is_null())
             {
                 etcd_alive++;
-                etcd_db_size = etcd_states[i]["dbSizeInUse"].uint64_value();
+                etcd_db_size = etcd_states[i]["dbSize"].uint64_value();
             }
         }
         int mon_count = 0;
@@ -132,7 +132,7 @@ resume_2:
             auto kv = parent->cli->st_cli.parse_etcd_kv(osd_stats[i]);
             osd_num_t stat_osd_num = 0;
             char null_byte = 0;
-            int scanned = sscanf(kv.key.c_str() + parent->cli->st_cli.etcd_prefix.size(), "/osd/stats/%lu%c", &stat_osd_num, &null_byte);
+            int scanned = sscanf(kv.key.c_str() + parent->cli->st_cli.etcd_prefix.size(), "/osd/stats/%ju%c", &stat_osd_num, &null_byte);
             if (scanned != 1 || !stat_osd_num)
             {
                 fprintf(stderr, "Invalid key in etcd: %s\n", kv.key.c_str());
@@ -283,7 +283,7 @@ resume_2:
         }
         printf(
             "  cluster:\n"
-            "    etcd: %d / %ld up, %s database size\n"
+            "    etcd: %d / %zd up, %s database size\n"
             "    mon:  %d up%s\n"
             "    osd:  %d / %d up\n"
             "  \n"

src/cluster_client.cpp

@@ -265,7 +265,7 @@ void cluster_client_t::erase_op(cluster_op_t *op)
     }
 }
 
-void cluster_client_t::continue_ops(bool up_retry)
+void cluster_client_t::continue_ops(int time_passed)
 {
     if (!pgs_loaded)
     {
@@ -277,22 +277,27 @@ void cluster_client_t::continue_ops(bool up_retry)
         // Attempt to reenter the function
         return;
     }
+    int reset_duration = 0;
 restart:
     continuing_ops = 1;
     for (auto op = op_queue_head; op; )
     {
         cluster_op_t *next_op = op->next;
-        if (!op->up_wait || up_retry)
+        if (op->retry_after && time_passed)
         {
-            op->up_wait = false;
+            op->retry_after = op->retry_after > time_passed ? op->retry_after-time_passed : 0;
-            if (!op->prev_wait)
+            if (op->retry_after && (!reset_duration || op->retry_after < reset_duration))
             {
-                if (op->opcode == OSD_OP_SYNC)
+                reset_duration = op->retry_after;
-                    continue_sync(op);
-                else
-                    continue_rw(op);
             }
         }
+        if (!op->retry_after && !op->prev_wait)
+        {
+            if (op->opcode == OSD_OP_SYNC)
+                continue_sync(op);
+            else
+                continue_rw(op);
+        }
         op = next_op;
         if (continuing_ops == 2)
         {
@@ -300,6 +305,27 @@ restart:
         }
     }
     continuing_ops = 0;
+    reset_retry_timer(reset_duration);
+}
+
+void cluster_client_t::reset_retry_timer(int new_duration)
+{
+    if (retry_timeout_duration && retry_timeout_duration <= new_duration || !new_duration)
+    {
+        return;
+    }
+    if (retry_timeout_id)
+    {
+        tfd->clear_timer(retry_timeout_id);
+    }
+    retry_timeout_duration = new_duration;
+    retry_timeout_id = tfd->set_timer(retry_timeout_duration, false, [this](int)
+    {
+        int time_passed = retry_timeout_duration;
+        retry_timeout_id = 0;
+        retry_timeout_duration = 0;
+        continue_ops(time_passed);
+    });
 }
 
 void cluster_client_t::on_load_config_hook(json11::Json::object & etcd_global_config)
@@ -349,15 +375,25 @@ void cluster_client_t::on_load_config_hook(json11::Json::object & etcd_global_co
     {
         client_max_writeback_iodepth = DEFAULT_CLIENT_MAX_WRITEBACK_IODEPTH;
     }
-    // up_wait_retry_interval
+    // client_retry_interval
-    up_wait_retry_interval = config["up_wait_retry_interval"].uint64_value();
+    client_retry_interval = config["client_retry_interval"].uint64_value();
-    if (!up_wait_retry_interval)
+    if (!client_retry_interval)
     {
-        up_wait_retry_interval = 50;
+        client_retry_interval = 50;
     }
-    else if (up_wait_retry_interval < 10)
+    else if (client_retry_interval < 10)
     {
-        up_wait_retry_interval = 10;
+        client_retry_interval = 10;
+    }
+    // client_eio_retry_interval
+    client_eio_retry_interval = 1000;
+    if (!config["client_eio_retry_interval"].is_null())
+    {
+        client_eio_retry_interval = config["client_eio_retry_interval"].uint64_value();
+        if (client_eio_retry_interval && client_eio_retry_interval < 10)
+        {
+            client_eio_retry_interval = 10;
+        }
     }
     // log_level
     log_level = config["log_level"].uint64_value();
@@ -716,15 +752,8 @@ resume_1:
                 // We'll need to retry again
                 if (op->parts[i].flags & PART_RETRY)
                 {
-                    op->up_wait = true;
+                    op->retry_after = client_retry_interval;
-                    if (!retry_timeout_id)
+                    reset_retry_timer(client_retry_interval);
-                    {
-                        retry_timeout_id = tfd->set_timer(up_wait_retry_interval, false, [this](int)
-                        {
-                            retry_timeout_id = 0;
-                            continue_ops(true);
-                        });
-                    }
                 }
                 op->state = 1;
             }
@@ -780,10 +809,9 @@ resume_2:
         return 1;
     }
     else if (op->retval != 0 && !(op->flags & OP_FLUSH_BUFFER) &&
-        op->retval != -EPIPE && op->retval != -EIO && op->retval != -ENOSPC)
+        op->retval != -EPIPE && (op->retval != -EIO || !client_eio_retry_interval) && op->retval != -ENOSPC)
     {
         // Fatal error (neither -EPIPE, -EIO nor -ENOSPC)
-        // FIXME: Add a parameter to allow to not wait for EIOs (incomplete or corrupted objects) to heal
         erase_op(op);
         return 1;
     }
@@ -1156,7 +1184,7 @@ void cluster_client_t::handle_op_part(cluster_op_part_t *part)
             if (op->retval != -EPIPE || log_level > 0)
             {
                 fprintf(
-                    stderr, "%s operation failed on OSD %lu: retval=%ld (expected %d), dropping connection\n",
+                    stderr, "%s operation failed on OSD %ju: retval=%jd (expected %d), dropping connection\n",
                     osd_op_names[part->op.req.hdr.opcode], part->osd_num, part->op.reply.hdr.retval, expected
                 );
             }
@@ -1164,23 +1192,19 @@ void cluster_client_t::handle_op_part(cluster_op_part_t *part)
         else
         {
             fprintf(
-                stderr, "%s operation failed on OSD %lu: retval=%ld (expected %d)\n",
+                stderr, "%s operation failed on OSD %ju: retval=%jd (expected %d)\n",
                 osd_op_names[part->op.req.hdr.opcode], part->osd_num, part->op.reply.hdr.retval, expected
             );
         }
         // All next things like timer, continue_sync/rw and stop_client may affect the operation again
         // So do all these things after modifying operation state, otherwise we may hit reenterability bugs
         // FIXME postpone such things to set_immediate here to avoid bugs
-        // Mark op->up_wait = true to retry operation after a short pause (not immediately)
+        // Set op->retry_after to retry operation after a short pause (not immediately)
-        op->up_wait = true;
+        if (!op->retry_after)
-        if (!retry_timeout_id)
         {
-            retry_timeout_id = tfd->set_timer(up_wait_retry_interval, false, [this](int)
+            op->retry_after = op->retval == -EIO ? client_eio_retry_interval : client_retry_interval;
-            {
-                retry_timeout_id = 0;
-                continue_ops(true);
-            });
         }
+        reset_retry_timer(op->retry_after);
         if (op->inflight_count == 0)
         {
             if (op->opcode == OSD_OP_SYNC)

src/cluster_client.h

@@ -59,7 +59,7 @@ protected:
     void *buf = NULL;
     cluster_op_t *orig_op = NULL;
     bool needs_reslice = false;
-    bool up_wait = false;
+    int retry_after = 0;
     int inflight_count = 0, done_count = 0;
     std::vector<cluster_op_part_t> parts;
     void *part_bitmaps = NULL;
@@ -92,9 +92,11 @@ class cluster_client_t
     uint64_t client_max_writeback_iodepth = 0;
 
     int log_level = 0;
-    int up_wait_retry_interval = 500; // ms
+    int client_retry_interval = 50; // ms
+    int client_eio_retry_interval = 1000; // ms
 
     int retry_timeout_id = 0;
+    int retry_timeout_duration = 0;
     std::vector<cluster_op_t*> offline_ops;
     cluster_op_t *op_queue_head = NULL, *op_queue_tail = NULL;
     writeback_cache_t *wb = NULL;
@@ -131,7 +133,7 @@ public:
 
     bool get_immediate_commit(uint64_t inode);
 
-    void continue_ops(bool up_retry = false);
+    void continue_ops(int time_passed = 0);
     inode_list_t *list_inode_start(inode_t inode,
         std::function<void(inode_list_t* lst, std::set<object_id>&& objects, pg_num_t pg_num, osd_num_t primary_osd, int status)> callback);
     int list_pg_count(inode_list_t *lst);
@@ -152,6 +154,7 @@ protected:
     int continue_rw(cluster_op_t *op);
     bool check_rw(cluster_op_t *op);
     void slice_rw(cluster_op_t *op);
+    void reset_retry_timer(int new_duration);
     bool try_send(cluster_op_t *op, int i);
     int continue_sync(cluster_op_t *op);
     void send_sync(cluster_op_t *op, cluster_op_part_t *part);

src/cluster_client_list.cpp

@@ -226,7 +226,7 @@ void cluster_client_t::send_list(inode_list_osd_t *cur_list)
     {
         if (op->reply.hdr.retval < 0)
         {
-            fprintf(stderr, "Failed to get PG %u/%u object list from OSD %lu (retval=%ld), skipping\n",
+            fprintf(stderr, "Failed to get PG %u/%u object list from OSD %ju (retval=%jd), skipping\n",
                 cur_list->pg->lst->pool_id, cur_list->pg->pg_num, cur_list->osd_num, op->reply.hdr.retval);
         }
         else
@@ -236,7 +236,7 @@ void cluster_client_t::send_list(inode_list_osd_t *cur_list)
                 // Unstable objects, if present, mean that someone still writes into the inode. Warn the user about it.
                 cur_list->pg->has_unstable = true;
                 fprintf(
-                    stderr, "[PG %u/%u] Inode still has %lu unstable object versions out of total %lu - is it still open?\n",
+                    stderr, "[PG %u/%u] Inode still has %ju unstable object versions out of total %ju - is it still open?\n",
                     cur_list->pg->lst->pool_id, cur_list->pg->pg_num, op->reply.hdr.retval - op->reply.sec_list.stable_count,
                     op->reply.hdr.retval
                 );
@@ -244,7 +244,7 @@ void cluster_client_t::send_list(inode_list_osd_t *cur_list)
             if (log_level > 0)
             {
                 fprintf(
-                    stderr, "[PG %u/%u] Got inode object list from OSD %lu: %ld object versions\n",
+                    stderr, "[PG %u/%u] Got inode object list from OSD %ju: %jd object versions\n",
                     cur_list->pg->lst->pool_id, cur_list->pg->pg_num, cur_list->osd_num, op->reply.hdr.retval
                 );
             }

src/disk_simple_offsets.cpp

@@ -47,7 +47,7 @@ void disk_tool_simple_offsets(json11::Json cfg, bool json_output)
     if (!bitmap_granularity)
         bitmap_granularity = DEFAULT_BITMAP_GRANULARITY;
     if (!journal_size)
-        journal_size = 16*1024*1024;
+        journal_size = 32*1024*1024;
     if (!device_block_size)
         device_block_size = 4096;
     if (!data_csum_type)
@@ -75,9 +75,9 @@ void disk_tool_simple_offsets(json11::Json cfg, bool json_output)
             if (st.st_blksize < device_block_size)
             {
                 fprintf(
-                    stderr, "Warning: %s reports %lu byte blocks, but we use %lu."
+                    stderr, "Warning: %s reports %ju byte blocks, but we use %ju."
-                    " Set --device_block_size=%lu if you're sure it works well with %lu byte blocks.\n",
+                    " Set --device_block_size=%ju if you're sure it works well with %ju byte blocks.\n",
-                    device.c_str(), st.st_blksize, device_block_size, st.st_blksize, st.st_blksize
+                    device.c_str(), (uint64_t)st.st_blksize, device_block_size, (uint64_t)st.st_blksize, (uint64_t)st.st_blksize
                 );
             }
         }
@@ -99,19 +99,19 @@ void disk_tool_simple_offsets(json11::Json cfg, bool json_output)
     if (device_block_size < 512 || device_block_size > 1048576 ||
         device_block_size & (device_block_size-1) != 0)
     {
-        fprintf(stderr, "Invalid device block size specified: %lu\n", device_block_size);
+        fprintf(stderr, "Invalid device block size specified: %ju\n", device_block_size);
         exit(1);
     }
     if (data_block_size < device_block_size || data_block_size > MAX_DATA_BLOCK_SIZE ||
         data_block_size & (data_block_size-1) != 0)
     {
-        fprintf(stderr, "Invalid object size specified: %lu\n", data_block_size);
+        fprintf(stderr, "Invalid object size specified: %ju\n", data_block_size);
         exit(1);
     }
     if (bitmap_granularity < device_block_size || bitmap_granularity > data_block_size ||
         bitmap_granularity & (bitmap_granularity-1) != 0)
     {
-        fprintf(stderr, "Invalid bitmap granularity specified: %lu\n", bitmap_granularity);
+        fprintf(stderr, "Invalid bitmap granularity specified: %ju\n", bitmap_granularity);
         exit(1);
     }
     if (csum_block_size && (data_block_size % csum_block_size))
@@ -145,8 +145,8 @@ void disk_tool_simple_offsets(json11::Json cfg, bool json_output)
     {
         // Env
         printf(
-            "meta_block_size=%lu\njournal_block_size=%lu\ndata_size=%lu\n"
+            "meta_block_size=%ju\njournal_block_size=%ju\ndata_size=%ju\n"
-            "data_device=%s\njournal_offset=%lu\nmeta_offset=%lu\ndata_offset=%lu\n",
+            "data_device=%s\njournal_offset=%ju\nmeta_offset=%ju\ndata_offset=%ju\n",
             device_block_size, device_block_size, device_size-data_offset,
             device.c_str(), journal_offset, meta_offset, data_offset
         );
@@ -160,14 +160,14 @@ void disk_tool_simple_offsets(json11::Json cfg, bool json_output)
         }
         if (device_block_size != 4096)
         {
-            printf("--meta_block_size %lu\n--journal_block_size %lu\n", device_block_size, device_block_size);
+            printf("--meta_block_size %ju\n--journal_block_size %ju\n", device_block_size, device_block_size);
         }
         if (orig_device_size)
         {
-            printf("--data_size %lu\n", device_size-data_offset);
+            printf("--data_size %ju\n", device_size-data_offset);
         }
         printf(
-            "--data_device %s\n--journal_offset %lu\n--meta_offset %lu\n--data_offset %lu\n",
+            "--data_device %s\n--journal_offset %ju\n--meta_offset %ju\n--data_offset %ju\n",
             device.c_str(), journal_offset, meta_offset, data_offset
         );
     }

src/disk_tool.cpp

@@ -167,7 +167,7 @@ static const char *help_text =
     "  Calculate offsets for old simple&stupid (no superblock) OSD deployment. Options:\n"
     "    --object_size 128k       Set blockstore block size\n"
     "    --bitmap_granularity 4k  Set bitmap granularity\n"
-    "    --journal_size 16M       Set journal size\n"
+    "    --journal_size 32M       Set journal size\n"
     "    --data_csum_type none    Set data checksum type (crc32c or none)\n"
     "    --csum_block_size 4k     Set data checksum block size\n"
     "    --device_block_size 4k   Set device block size\n"

src/disk_tool.h

@@ -4,7 +4,7 @@
 #pragma once
 
 #ifndef _LARGEFILE64_SOURCE
-#define _LARGEFILE64_SOURCE 1
+#define _LARGEFILE64_SOURCE
 #endif
 
 #include <map>

src/disk_tool_journal.cpp

@@ -38,7 +38,7 @@ int disk_tool_t::dump_journal()
             }
             if (json)
             {
-                printf("%s{\"offset\":\"0x%lx\"", first_block ? "" : ",\n", journal_pos);
+                printf("%s{\"offset\":\"0x%jx\"", first_block ? "" : ",\n", journal_pos);
                 first_block = false;
             }
             if (s == dsk.journal_block_size)
@@ -46,13 +46,13 @@ int disk_tool_t::dump_journal()
                 if (json)
                     printf(",\"type\":\"zero\"}");
                 else
-                    printf("offset %08lx: zeroes\n", journal_pos);
+                    printf("offset %08jx: zeroes\n", journal_pos);
                 journal_pos += dsk.journal_block_size;
             }
             else if (((journal_entry*)journal_buf)->magic == JOURNAL_MAGIC)
             {
                 if (!json)
-                    printf("offset %08lx:\n", journal_pos);
+                    printf("offset %08jx:\n", journal_pos);
                 else
                     printf(",\"entries\":[\n");
                 if (journal_pos == 0)
@@ -80,9 +80,9 @@ int disk_tool_t::dump_journal()
             else
             {
                 if (json)
-                    printf(",\"type\":\"data\",\"pattern\":\"%08lx\"}", *((uint64_t*)journal_buf));
+                    printf(",\"type\":\"data\",\"pattern\":\"%08jx\"}", *((uint64_t*)journal_buf));
                 else
-                    printf("offset %08lx: no magic in the beginning, looks like random data (pattern=%08lx)\n", journal_pos, *((uint64_t*)journal_buf));
+                    printf("offset %08jx: no magic in the beginning, looks like random data (pattern=%08jx)\n", journal_pos, *((uint64_t*)journal_buf));
                 journal_pos += dsk.journal_block_size;
             }
         }
@@ -98,12 +98,12 @@ int disk_tool_t::dump_journal()
             if (json && dump_with_blocks)
                 first_entry = true;
             if (!json)
-                printf("offset %08lx:\n", journal_pos);
+                printf("offset %08jx:\n", journal_pos);
             auto pos = journal_pos;
             int r = process_journal_block(data, [this, pos](int num, journal_entry *je)
             {
                 if (json && dump_with_blocks && first_entry)
-                    printf("%s{\"offset\":\"0x%lx\",\"entries\":[\n", first_block ? "" : ",\n", pos);
+                    printf("%s{\"offset\":\"0x%jx\",\"entries\":[\n", first_block ? "" : ",\n", pos);
                 dump_journal_entry(num, je, json);
                 first_block = false;
             });
@@ -134,12 +134,12 @@ int disk_tool_t::process_journal(std::function<int(void*)> block_fn)
     journal_entry *je = (journal_entry*)(data);
     if (je->magic != JOURNAL_MAGIC || je->type != JE_START || je_crc32(je) != je->crc32)
     {
-        fprintf(stderr, "offset %08lx: journal superblock is invalid\n", journal_pos);
+        fprintf(stderr, "offset %08jx: journal superblock is invalid\n", journal_pos);
         r = 1;
     }
     else if (je->start.size != JE_START_V0_SIZE && je->start.version != JOURNAL_VERSION_V1 && je->start.version != JOURNAL_VERSION_V2)
     {
-        fprintf(stderr, "offset %08lx: journal superblock contains version %lu, but I only understand 0, 1 and 2\n",
+        fprintf(stderr, "offset %08jx: journal superblock contains version %ju, but I only understand 0, 1 and 2\n",
             journal_pos, je->start.size == JE_START_V0_SIZE ? 0 : je->start.version);
         r = 1;
     }
@@ -296,7 +296,7 @@ void disk_tool_t::dump_journal_entry(int num, journal_entry *je, bool json)
     if (je->type == JE_START)
     {
         printf(
-            json ? ",\"type\":\"start\",\"start\":\"0x%lx\"" : "je_start start=%08lx",
+            json ? ",\"type\":\"start\",\"start\":\"0x%jx\"" : "je_start start=%08jx",
             je->start.journal_start
         );
         if (je->start.data_csum_type)
@@ -312,15 +312,15 @@ void disk_tool_t::dump_journal_entry(int num, journal_entry *je, bool json)
     {
         auto & sw = je->small_write;
         printf(
-            json ? ",\"type\":\"small_write%s\",\"inode\":\"0x%lx\",\"stripe\":\"0x%lx\",\"ver\":\"%lu\",\"offset\":%u,\"len\":%u,\"loc\":\"0x%lx\""
+            json ? ",\"type\":\"small_write%s\",\"inode\":\"0x%jx\",\"stripe\":\"0x%jx\",\"ver\":\"%ju\",\"offset\":%u,\"len\":%u,\"loc\":\"0x%jx\""
-                : "je_small_write%s oid=%lx:%lx ver=%lu offset=%u len=%u loc=%08lx",
+                : "je_small_write%s oid=%jx:%jx ver=%ju offset=%u len=%u loc=%08jx",
             je->type == JE_SMALL_WRITE_INSTANT ? "_instant" : "",
             sw.oid.inode, sw.oid.stripe, sw.version, sw.offset, sw.len, sw.data_offset
         );
         if (journal_calc_data_pos != sw.data_offset)
         {
-            printf(json ? ",\"bad_loc\":true,\"calc_loc\":\"0x%lx\""
+            printf(json ? ",\"bad_loc\":true,\"calc_loc\":\"0x%jx\""
-                : " (mismatched, calculated = %08lx)", journal_pos);
+                : " (mismatched, calculated = %08jx)", journal_pos);
         }
         uint32_t data_csum_size = (!je_start.csum_block_size
             ? 0
@@ -367,8 +367,8 @@ void disk_tool_t::dump_journal_entry(int num, journal_entry *je, bool json)
     {
         auto & bw = je->big_write;
         printf(
-            json ? ",\"type\":\"big_write%s\",\"inode\":\"0x%lx\",\"stripe\":\"0x%lx\",\"ver\":\"%lu\",\"offset\":%u,\"len\":%u,\"loc\":\"0x%lx\""
+            json ? ",\"type\":\"big_write%s\",\"inode\":\"0x%jx\",\"stripe\":\"0x%jx\",\"ver\":\"%ju\",\"offset\":%u,\"len\":%u,\"loc\":\"0x%jx\""
-                : "je_big_write%s oid=%lx:%lx ver=%lu offset=%u len=%u loc=%08lx",
+                : "je_big_write%s oid=%jx:%jx ver=%ju offset=%u len=%u loc=%08jx",
             je->type == JE_BIG_WRITE_INSTANT ? "_instant" : "",
             bw.oid.inode, bw.oid.stripe, bw.version, bw.offset, bw.len, bw.location
         );
@@ -398,24 +398,24 @@ void disk_tool_t::dump_journal_entry(int num, journal_entry *je, bool json)
     else if (je->type == JE_STABLE)
     {
         printf(
-            json ? ",\"type\":\"stable\",\"inode\":\"0x%lx\",\"stripe\":\"0x%lx\",\"ver\":\"%lu\"}"
+            json ? ",\"type\":\"stable\",\"inode\":\"0x%jx\",\"stripe\":\"0x%jx\",\"ver\":\"%ju\"}"
-                : "je_stable oid=%lx:%lx ver=%lu\n",
+                : "je_stable oid=%jx:%jx ver=%ju\n",
             je->stable.oid.inode, je->stable.oid.stripe, je->stable.version
         );
     }
     else if (je->type == JE_ROLLBACK)
     {
         printf(
-            json ? ",\"type\":\"rollback\",\"inode\":\"0x%lx\",\"stripe\":\"0x%lx\",\"ver\":\"%lu\"}"
+            json ? ",\"type\":\"rollback\",\"inode\":\"0x%jx\",\"stripe\":\"0x%jx\",\"ver\":\"%ju\"}"
-                : "je_rollback oid=%lx:%lx ver=%lu\n",
+                : "je_rollback oid=%jx:%jx ver=%ju\n",
             je->rollback.oid.inode, je->rollback.oid.stripe, je->rollback.version
         );
     }
     else if (je->type == JE_DELETE)
     {
         printf(
-            json ? ",\"type\":\"delete\",\"inode\":\"0x%lx\",\"stripe\":\"0x%lx\",\"ver\":\"%lu\"}"
+            json ? ",\"type\":\"delete\",\"inode\":\"0x%jx\",\"stripe\":\"0x%jx\",\"ver\":\"%ju\"}"
-                : "je_delete oid=%lx:%lx ver=%lu\n",
+                : "je_delete oid=%jx:%jx ver=%ju\n",
             je->del.oid.inode, je->del.oid.stripe, je->del.version
         );
     }

src/disk_tool_meta.cpp

@@ -54,7 +54,7 @@ int disk_tool_t::process_meta(std::function<void(blockstore_meta_header_v2_t *)>
         else
         {
             // Unsupported version
-            fprintf(stderr, "Metadata format is too new for me (stored version is %lu, max supported %u).\n", hdr->version, BLOCKSTORE_META_FORMAT_V2);
+            fprintf(stderr, "Metadata format is too new for me (stored version is %ju, max supported %u).\n", hdr->version, BLOCKSTORE_META_FORMAT_V2);
             free(data);
             close(dsk.meta_fd);
             dsk.meta_fd = -1;
@@ -108,7 +108,7 @@ int disk_tool_t::process_meta(std::function<void(blockstore_meta_header_v2_t *)>
                             uint32_t *entry_csum = (uint32_t*)((uint8_t*)entry + dsk.clean_entry_size - 4);
                             if (*entry_csum != crc32c(0, entry, dsk.clean_entry_size - 4))
                             {
-                                fprintf(stderr, "Metadata entry %lu is corrupt (checksum mismatch), skipping\n", block_num);
+                                fprintf(stderr, "Metadata entry %ju is corrupt (checksum mismatch), skipping\n", block_num);
                                 continue;
                             }
                         }
@@ -184,7 +184,7 @@ void disk_tool_t::dump_meta_header(blockstore_meta_header_v2_t *hdr)
     }
     else
     {
-        printf("{\"version\":\"0.5\",\"meta_block_size\":%lu,\"entries\":[\n", dsk.meta_block_size);
+        printf("{\"version\":\"0.5\",\"meta_block_size\":%ju,\"entries\":[\n", dsk.meta_block_size);
     }
     first_entry = true;
 }
@@ -192,7 +192,7 @@ void disk_tool_t::dump_meta_header(blockstore_meta_header_v2_t *hdr)
 void disk_tool_t::dump_meta_entry(uint64_t block_num, clean_disk_entry *entry, uint8_t *bitmap)
 {
     printf(
-#define ENTRY_FMT "{\"block\":%lu,\"pool\":%u,\"inode\":\"0x%lx\",\"stripe\":\"0x%lx\",\"version\":%lu"
+#define ENTRY_FMT "{\"block\":%ju,\"pool\":%u,\"inode\":\"0x%jx\",\"stripe\":\"0x%jx\",\"version\":%ju"
         (first_entry ? ENTRY_FMT : (",\n" ENTRY_FMT)),
 #undef ENTRY_FMT
         block_num, INODE_POOL(entry->oid.inode), INODE_NO_POOL(entry->oid.inode),
@@ -265,7 +265,7 @@ int disk_tool_t::write_json_meta(json11::Json meta)
         {
             free(new_meta_buf);
             new_meta_buf = NULL;
-            fprintf(stderr, "Metadata (data block %lu) doesn't fit into the new area\n", data_block);
+            fprintf(stderr, "Metadata (data block %ju) doesn't fit into the new area\n", data_block);
             return 1;
         }
         clean_disk_entry *new_entry = (clean_disk_entry*)(new_meta_buf +

src/disk_tool_prepare.cpp

@@ -8,6 +8,7 @@
 int disk_tool_t::prepare_one(std::map<std::string, std::string> options, int is_hdd)
 {
     static const char *allow_additional_params[] = {
+        "autosync_writes",
         "data_io",
         "meta_io",
         "journal_io",
@@ -99,12 +100,9 @@ int disk_tool_t::prepare_one(std::map<std::string, std::string> options, int is_
         options["disable_journal_fsync"] = options["disable_data_fsync"];
     }
     // Calculate offsets if the same device is used for two or more of data, meta, and journal
-    if (options["journal_size"] == "")
+    if (options["journal_size"] == "" && (options["journal_device"] == "" || options["journal_device"] == options["data_device"]))
     {
-        if (options["journal_device"] == "")
+        options["journal_size"] = is_hdd || !json_is_true(options["disable_data_fsync"]) ? "128M" : "32M";
-            options["journal_size"] = is_hdd ? "128M" : "32M";
-        else if (is_hdd)
-            options["journal_size"] = DEFAULT_HYBRID_JOURNAL;
     }
     bool is_hybrid = is_hdd && options["journal_device"] != "" && options["journal_device"] != options["data_device"];
     if (is_hdd)
@@ -114,6 +112,15 @@ int disk_tool_t::prepare_one(std::map<std::string, std::string> options, int is_
         if (is_hybrid && options["throttle_small_writes"] == "")
             options["throttle_small_writes"] = "1";
     }
+    else if (!json_is_true(options["disable_data_fsync"]))
+    {
+        if (options.find("min_flusher_count") == options.end())
+            options["min_flusher_count"] = "32";
+        if (options.find("max_flusher_count") == options.end())
+            options["max_flusher_count"] = "256";
+        if (options.find("autosync_writes") == options.end())
+            options["autosync_writes"] = "512";
+    }
     json11::Json::object sb;
     blockstore_disk_t dsk;
     try
@@ -203,10 +210,10 @@ int disk_tool_t::prepare_one(std::map<std::string, std::string> options, int is_
         desc += " with metadata on "+realpath_str(options["meta_device"]);
     if (sep_j)
         desc += (sep_m ? " and journal on " : " with journal on ") + realpath_str(options["journal_device"]);
-    fprintf(stderr, "Initialized OSD %lu on %s\n", osd_num, desc.c_str());
+    fprintf(stderr, "Initialized OSD %ju on %s\n", osd_num, desc.c_str());
     if (shell_exec({ "systemctl", "enable", "--now", "vitastor-osd@"+std::to_string(osd_num) }, "", NULL, NULL) != 0)
     {
-        fprintf(stderr, "Failed to enable systemd unit vitastor-osd@%lu\n", osd_num);
+        fprintf(stderr, "Failed to enable systemd unit vitastor-osd@%ju\n", osd_num);
         return 1;
     }
     return 0;
@@ -330,7 +337,7 @@ json11::Json disk_tool_t::add_partitions(vitastor_dev_info_t & devinfo, std::vec
     std::string out;
     if (shell_exec({ "sfdisk", "--no-reread", "--force", devinfo.path }, script, &out, NULL) != 0)
     {
-        fprintf(stderr, "Failed to add %lu partition(s) with sfdisk\n", sizes.size());
+        fprintf(stderr, "Failed to add %zu partition(s) with sfdisk\n", sizes.size());
         return {};
     }
     // Get new partition table and find created partitions
@@ -345,7 +352,7 @@ json11::Json disk_tool_t::add_partitions(vitastor_dev_info_t & devinfo, std::vec
     }
     if (new_parts.size() != sizes.size())
     {
-        fprintf(stderr, "Failed to add %lu partition(s) with sfdisk: new partitions not found in table\n", sizes.size());
+        fprintf(stderr, "Failed to add %zu partition(s) with sfdisk: new partitions not found in table\n", sizes.size());
         return {};
     }
     // Check if new nodes exist and run partprobe if not
@@ -449,7 +456,7 @@ std::vector<std::string> disk_tool_t::get_new_data_parts(vitastor_dev_info_t & d
                     bool is_journal = sb["params"]["journal_device"].string_value() == part_path;
                     bool is_data = sb["params"]["data_device"].string_value() == part_path;
                     fprintf(
-                        stderr, "%s is already initialized for OSD %lu%s, skipping\n",
+                        stderr, "%s is already initialized for OSD %ju%s, skipping\n",
                         part["node"].string_value().c_str(), sb["params"]["osd_num"].uint64_value(),
                         (is_data ? " data" : (is_meta ? " meta" : (is_journal ? " journal" : "")))
                     );
@@ -532,7 +539,7 @@ int disk_tool_t::get_meta_partition(std::vector<vitastor_dev_info_t> & ssds, std
     if (sel < 0)
     {
         fprintf(
-            stderr, "Could not find free space for new SSD journal and metadata (need %lu + %lu MiB)\n",
+            stderr, "Could not find free space for new SSD journal and metadata (need %ju + %ju MiB)\n",
             meta_size/1024/1024, journal_size/1024/1024
         );
         return 1;
@@ -616,6 +623,7 @@ int disk_tool_t::prepare(std::vector<std::string> devices)
         options.erase("disable_meta_fsync");
         options.erase("disable_journal_fsync");
     }
+    auto journal_size = options["journal_size"];
     for (auto & dev: devinfo)
     {
         if (!hybrid || dev.is_hdd)
@@ -633,11 +641,13 @@ int disk_tool_t::prepare(std::vector<std::string> devices)
                     {
                         return 1;
                     }
+                    options.erase("journal_size");
                 }
                 // Treat all disks as SSDs if not in the hybrid mode
                 prepare_one(options, dev.is_hdd ? 1 : 0);
                 if (hybrid)
                 {
+                    options["journal_size"] = journal_size;
                     options.erase("journal_device");
                     options.erase("meta_device");
                 }

src/disk_tool_resize.cpp

@@ -184,7 +184,7 @@ void disk_tool_t::resize_init(blockstore_meta_header_v2_t *hdr)
     }
     if (new_meta_len < dsk.meta_block_size*new_meta_blocks)
     {
-        fprintf(stderr, "New metadata area size is too small, should be at least %lu bytes\n", dsk.meta_block_size*new_meta_blocks);
+        fprintf(stderr, "New metadata area size is too small, should be at least %ju bytes\n", dsk.meta_block_size*new_meta_blocks);
         exit(1);
     }
     // Check that new metadata, journal and data areas don't overlap
@@ -289,7 +289,7 @@ int disk_tool_t::resize_copy_data()
                         if (data->res != dsk.data_block_size)
                         {
                             fprintf(
-                                stderr, "Failed to read %u bytes at %lu from %s: %s\n", dsk.data_block_size,
+                                stderr, "Failed to read %u bytes at %ju from %s: %s\n", dsk.data_block_size,
                                 dsk.data_offset + moving_blocks[i].old_loc*dsk.data_block_size, dsk.data_device.c_str(),
                                 data->res < 0 ? strerror(-data->res) : "short read"
                             );
@@ -314,7 +314,7 @@ int disk_tool_t::resize_copy_data()
                         if (data->res != dsk.data_block_size)
                         {
                             fprintf(
-                                stderr, "Failed to write %u bytes at %lu to %s: %s\n", dsk.data_block_size,
+                                stderr, "Failed to write %u bytes at %ju to %s: %s\n", dsk.data_block_size,
                                 dsk.data_offset + moving_blocks[i].new_loc*dsk.data_block_size, dsk.data_device.c_str(),
                                 data->res < 0 ? strerror(-data->res) : "short write"
                             );

src/disk_tool_udev.cpp

@@ -43,8 +43,8 @@ int disk_tool_t::udev_import(std::string device)
     }
     uint64_t osd_num = sb["params"]["osd_num"].uint64_value();
     // Print variables for udev
-    printf("VITASTOR_OSD_NUM=%lu\n", osd_num);
+    printf("VITASTOR_OSD_NUM=%ju\n", osd_num);
-    printf("VITASTOR_ALIAS=osd%lu-%s\n", osd_num, sb["device_type"].string_value().c_str());
+    printf("VITASTOR_ALIAS=osd%ju-%s\n", osd_num, sb["device_type"].string_value().c_str());
     printf("VITASTOR_DATA_DEVICE=%s\n", udev_escape(sb["params"]["data_device"].string_value()).c_str());
     if (sb["real_meta_device"].string_value() != "" && sb["real_meta_device"] != sb["real_data_device"])
         printf("VITASTOR_META_DEVICE=%s\n", udev_escape(sb["params"]["meta_device"].string_value()).c_str());
@@ -466,12 +466,12 @@ int disk_tool_t::purge_devices(const std::vector<std::string> & devices)
                     close(fd);
                 if (r != 0)
                 {
-                    fprintf(stderr, "Failed to clear OSD %lu %s device %s superblock: %s\n",
+                    fprintf(stderr, "Failed to clear OSD %ju %s device %s superblock: %s\n",
                         sb["params"]["osd_num"].uint64_value(), dev_type.c_str(), dev.c_str(), strerror(errno));
                 }
                 else
                 {
-                    fprintf(stderr, "OSD %lu %s device %s superblock cleared\n",
+                    fprintf(stderr, "OSD %ju %s device %s superblock cleared\n",
                         sb["params"]["osd_num"].uint64_value(), dev_type.c_str(), dev.c_str());
                 }
                 if (sb["params"][dev_type+"_device"].string_value().substr(0, 22) == "/dev/disk/by-partuuid/")

src/disk_tool_utils.cpp

@@ -12,9 +12,9 @@ uint64_t sscanf_json(const char *fmt, const json11::Json & str)
 {
     uint64_t value = 0;
     if (fmt)
-        sscanf(str.string_value().c_str(), "%lx", &value);
+        sscanf(str.string_value().c_str(), "%jx", &value);
     else if (str.string_value().size() > 2 && (str.string_value()[0] == '0' && str.string_value()[1] == 'x'))
-        sscanf(str.string_value().c_str(), "0x%lx", &value);
+        sscanf(str.string_value().c_str(), "0x%jx", &value);
     else
         value = str.uint64_value();
     return value;

src/etcd_state_client.cpp

@@ -333,7 +333,7 @@ void etcd_state_client_t::start_etcd_watcher()
         etcd_watch_ws = NULL;
     }
     if (this->log_level > 1)
-        fprintf(stderr, "Trying to connect to etcd websocket at %s, watch from revision %lu\n", etcd_address.c_str(), etcd_watch_revision);
+        fprintf(stderr, "Trying to connect to etcd websocket at %s, watch from revision %ju\n", etcd_address.c_str(), etcd_watch_revision);
     etcd_watch_ws = open_websocket(tfd, etcd_address, etcd_api_path+"/watch", etcd_slow_timeout,
         [this, cur_addr = selected_etcd_address](const http_response_t *msg)
     {
@@ -357,7 +357,7 @@ void etcd_state_client_t::start_etcd_watcher()
                         watch_id == ETCD_OSD_STATE_WATCH_ID)
                         etcd_watches_initialised++;
                     if (etcd_watches_initialised == ETCD_TOTAL_WATCHES && this->log_level > 0)
-                        fprintf(stderr, "Successfully subscribed to etcd at %s, revision %lu\n", cur_addr.c_str(), etcd_watch_revision);
+                        fprintf(stderr, "Successfully subscribed to etcd at %s, revision %ju\n", cur_addr.c_str(), etcd_watch_revision);
                 }
                 if (data["result"]["canceled"].bool_value())
                 {
@@ -371,7 +371,7 @@ void etcd_state_client_t::start_etcd_watcher()
                             // check to not trigger on_reload_hook multiple times
                             if (etcd_watch_ws != NULL)
                             {
-                                fprintf(stderr, "Revisions before %lu were compacted by etcd, reloading state\n",
+                                fprintf(stderr, "Revisions before %ju were compacted by etcd, reloading state\n",
                                     data["result"]["compact_revision"].uint64_value());
                                 http_close(etcd_watch_ws);
                                 etcd_watch_ws = NULL;
@@ -382,7 +382,7 @@ void etcd_state_client_t::start_etcd_watcher()
                         }
                         else
                         {
-                            fprintf(stderr, "Revisions before %lu were compacted by etcd, exiting\n",
+                            fprintf(stderr, "Revisions before %ju were compacted by etcd, exiting\n",
                                 data["result"]["compact_revision"].uint64_value());
                             exit(1);
                         }
@@ -646,7 +646,7 @@ void etcd_state_client_t::load_pgs()
             etcd_watch_revision = data["header"]["revision"].uint64_value()+1;
             if (this->log_level > 3)
             {
-                fprintf(stderr, "Loaded revision %lu of PG configuration\n", etcd_watch_revision-1);
+                fprintf(stderr, "Loaded revision %ju of PG configuration\n", etcd_watch_revision-1);
             }
         }
         for (auto & res: data["responses"].array_items())
@@ -740,7 +740,7 @@ void etcd_state_client_t::clean_nonexistent_pgs()
     {
         if (seen_peers.find(peer_item.first) == seen_peers.end())
         {
-            fprintf(stderr, "OSD %lu state disappeared after reload, forgetting it\n", peer_item.first);
+            fprintf(stderr, "OSD %ju state disappeared after reload, forgetting it\n", peer_item.first);
             parse_state((etcd_kv_t){
                 .key = etcd_prefix+"/osd/state/"+std::to_string(peer_item.first),
             });
@@ -890,7 +890,7 @@ void etcd_state_client_t::parse_state(const etcd_kv_t & kv)
             {
                 if (pg_item.second.target_set.size() != parsed_cfg.pg_size)
                 {
-                    fprintf(stderr, "Pool %u PG %u configuration is invalid: osd_set size %lu != pool pg_size %lu\n",
+                    fprintf(stderr, "Pool %u PG %u configuration is invalid: osd_set size %zu != pool pg_size %ju\n",
                         pool_id, pg_item.first, pg_item.second.target_set.size(), parsed_cfg.pg_size);
                     pg_item.second.pause = true;
                 }
@@ -936,7 +936,7 @@ void etcd_state_client_t::parse_state(const etcd_kv_t & kv)
                 }
                 if (parsed_cfg.target_set.size() != pool_config[pool_id].pg_size)
                 {
-                    fprintf(stderr, "Pool %u PG %u configuration is invalid: osd_set size %lu != pool pg_size %lu\n",
+                    fprintf(stderr, "Pool %u PG %u configuration is invalid: osd_set size %zu != pool pg_size %ju\n",
                         pool_id, pg_num, parsed_cfg.target_set.size(), pool_config[pool_id].pg_size);
                     parsed_cfg.pause = true;
                 }
@@ -950,7 +950,7 @@ void etcd_state_client_t::parse_state(const etcd_kv_t & kv)
                 if (pg_it->second.config_exists && pg_it->first != ++n)
                 {
                     fprintf(
-                        stderr, "Invalid pool %u PG configuration: PG numbers don't cover whole 1..%lu range\n",
+                        stderr, "Invalid pool %u PG configuration: PG numbers don't cover whole 1..%zu range\n",
                         pool_item.second.id, pool_item.second.pg_config.size()
                     );
                     for (pg_it = pool_item.second.pg_config.begin(); pg_it != pool_item.second.pg_config.end(); pg_it++)
@@ -1066,7 +1066,7 @@ void etcd_state_client_t::parse_state(const etcd_kv_t & kv)
                 (state & PG_PEERING) && state != PG_PEERING ||
                 (state & PG_INCOMPLETE) && state != PG_INCOMPLETE)
             {
-                fprintf(stderr, "Unexpected pool %u PG %u state in etcd: primary=%lu, state=%s\n", pool_id, pg_num, cur_primary, value["state"].dump().c_str());
+                fprintf(stderr, "Unexpected pool %u PG %u state in etcd: primary=%ju, state=%s\n", pool_id, pg_num, cur_primary, value["state"].dump().c_str());
                 return;
             }
             pg_cfg.cur_primary = cur_primary;
@@ -1102,7 +1102,7 @@ void etcd_state_client_t::parse_state(const etcd_kv_t & kv)
         uint64_t pool_id = 0;
         uint64_t inode_num = 0;
         char null_byte = 0;
-        int scanned = sscanf(key.c_str() + etcd_prefix.length()+14, "%lu/%lu%c", &pool_id, &inode_num, &null_byte);
+        int scanned = sscanf(key.c_str() + etcd_prefix.length()+14, "%ju/%ju%c", &pool_id, &inode_num, &null_byte);
         if (scanned != 2 || !pool_id || pool_id >= POOL_ID_MAX || !inode_num || (inode_num >> (64-POOL_ID_BITS)))
         {
             fprintf(stderr, "Bad etcd key %s, ignoring\n", key.c_str());
@@ -1145,7 +1145,7 @@ void etcd_state_client_t::parse_state(const etcd_kv_t & kv)
                     else if (parent_pool_id >= POOL_ID_MAX)
                     {
                         fprintf(
-                            stderr, "Inode %lu/%lu parent_pool value is invalid, ignoring parent setting\n",
+                            stderr, "Inode %ju/%ju parent_pool value is invalid, ignoring parent setting\n",
                             inode_num >> (64-POOL_ID_BITS), inode_num & (((uint64_t)1 << (64-POOL_ID_BITS)) - 1)
                         );
                         parent_inode_num = 0;

src/fio_cluster.cpp

@@ -377,7 +377,7 @@ static void io_callback(void *opaque, long retval)
     bsd->completed.push_back(io);
     if (bsd->trace)
     {
-        printf("--- %s 0x%lx retval=%ld\n", io->ddir == DDIR_READ ? "READ" :
+        printf("--- %s 0x%jx retval=%ld\n", io->ddir == DDIR_READ ? "READ" :
             (io->ddir == DDIR_WRITE ? "WRITE" : "SYNC"), (uint64_t)io, retval);
     }
 }
@@ -405,10 +405,11 @@ static enum fio_q_status sec_queue(struct thread_data *td, struct io_u *io)
     bsd->inflight++;
 
     uint64_t inode = opt->image ? vitastor_c_inode_get_num(bsd->watch) : opt->inode;
+    assert(io->xfer_buflen < (size_t)-1);
     switch (io->ddir)
     {
     case DDIR_READ:
-        iov = { .iov_base = io->xfer_buf, .iov_len = io->xfer_buflen };
+        iov = { .iov_base = io->xfer_buf, .iov_len = (size_t)io->xfer_buflen };
         vitastor_c_read(bsd->cli, inode, io->offset, io->xfer_buflen, &iov, 1, read_callback, io);
         bsd->last_sync = false;
         break;
@@ -436,7 +437,7 @@ static enum fio_q_status sec_queue(struct thread_data *td, struct io_u *io)
             io->error = EROFS;
             return FIO_Q_COMPLETED;
         }
-        iov = { .iov_base = io->xfer_buf, .iov_len = io->xfer_buflen };
+        iov = { .iov_base = io->xfer_buf, .iov_len = (size_t)io->xfer_buflen };
         vitastor_c_write(bsd->cli, inode, io->offset, io->xfer_buflen, 0, &iov, 1, io_callback, io);
         bsd->last_sync = false;
         break;
@@ -453,11 +454,11 @@ static enum fio_q_status sec_queue(struct thread_data *td, struct io_u *io)
     {
         if (io->ddir == DDIR_SYNC)
         {
-            printf("+++ SYNC 0x%lx\n", (uint64_t)io);
+            printf("+++ SYNC 0x%jx\n", (uint64_t)io);
         }
         else
         {
-            printf("+++ %s 0x%lx 0x%llx+%lx\n",
+            printf("+++ %s 0x%jx 0x%llx+%jx\n",
                 io->ddir == DDIR_READ ? "READ" : "WRITE",
                 (uint64_t)io, io->offset, (uint64_t)io->xfer_buflen);
         }

src/fio_sec_osd.cpp

@@ -310,7 +310,8 @@ static enum fio_q_status sec_queue(struct thread_data *td, struct io_u *io)
     int iovcnt = 1, wtotal = OSD_PACKET_SIZE;
     if (io->ddir == DDIR_WRITE)
     {
-        iov[iovcnt++] = { .iov_base = io->xfer_buf, .iov_len = io->xfer_buflen };
+        assert(io->xfer_buflen <= 0x7fffffff);
+        iov[iovcnt++] = { .iov_base = io->xfer_buf, .iov_len = (size_t)io->xfer_buflen };
         wtotal += io->xfer_buflen;
     }
     if (sendv_blocking(bsd->connect_fd, iov, iovcnt,
@@ -341,13 +342,13 @@ static int sec_getevents(struct thread_data *td, unsigned int min, unsigned int
         read_blocking(bsd->connect_fd, reply.buf, OSD_PACKET_SIZE);
         if (reply.hdr.magic != SECONDARY_OSD_REPLY_MAGIC)
         {
-            fprintf(stderr, "bad reply: magic = %lx instead of %lx\n", reply.hdr.magic, SECONDARY_OSD_REPLY_MAGIC);
+            fprintf(stderr, "bad reply: magic = %jx instead of %jx\n", reply.hdr.magic, SECONDARY_OSD_REPLY_MAGIC);
             exit(1);
         }
         auto it = bsd->queue.find(reply.hdr.id);
         if (it == bsd->queue.end())
         {
-            fprintf(stderr, "bad reply: op id %lx missing in local queue\n", reply.hdr.id);
+            fprintf(stderr, "bad reply: op id %jx missing in local queue\n", reply.hdr.id);
             exit(1);
         }
         io_u* io = it->second->fio_op;
@@ -357,7 +358,7 @@ static int sec_getevents(struct thread_data *td, unsigned int min, unsigned int
         {
             if (reply.hdr.retval != io->xfer_buflen)
             {
-                fprintf(stderr, "Short read: retval = %ld instead of %lu\n", reply.hdr.retval, (uint64_t)io->xfer_buflen);
+                fprintf(stderr, "Short read: retval = %jd instead of %ju\n", reply.hdr.retval, (uint64_t)io->xfer_buflen);
                 exit(1);
             }
             // Support bitmap
@@ -371,7 +372,8 @@ static int sec_getevents(struct thread_data *td, unsigned int min, unsigned int
                 else
                     iov[iovcnt++] = { .iov_base = (void*)(bitmap = (uint64_t)malloc(reply.sec_rw.attr_len)), .iov_len = reply.sec_rw.attr_len };
             }
-            iov[iovcnt++] = { .iov_base = io->xfer_buf, .iov_len = io->xfer_buflen };
+            assert(io->xfer_buflen <= 0x7FFFFFFF);
+            iov[iovcnt++] = { .iov_base = io->xfer_buf, .iov_len = (size_t)io->xfer_buflen };
             readv_blocking(bsd->connect_fd, iov, iovcnt);
             if (reply.sec_rw.attr_len > 8)
             {
@@ -382,7 +384,7 @@ static int sec_getevents(struct thread_data *td, unsigned int min, unsigned int
         {
             if (reply.hdr.retval != io->xfer_buflen)
             {
-                fprintf(stderr, "Short write: retval = %ld instead of %lu\n", reply.hdr.retval, (uint64_t)io->xfer_buflen);
+                fprintf(stderr, "Short write: retval = %jd instead of %ju\n", reply.hdr.retval, (uint64_t)io->xfer_buflen);
                 exit(1);
             }
         }
@@ -390,13 +392,13 @@ static int sec_getevents(struct thread_data *td, unsigned int min, unsigned int
         {
             if (reply.hdr.retval != 0)
             {
-                fprintf(stderr, "Sync failed: retval = %ld\n", reply.hdr.retval);
+                fprintf(stderr, "Sync failed: retval = %jd\n", reply.hdr.retval);
                 exit(1);
             }
         }
         if (opt->trace)
         {
-            printf("--- %s # %ld\n", io->ddir == DDIR_READ ? "READ" :
+            printf("--- %s # %ju\n", io->ddir == DDIR_READ ? "READ" :
                 (io->ddir == DDIR_WRITE ? "WRITE" : "SYNC"), reply.hdr.id);
         }
         bsd->completed.push_back(io);

src/malloc_or_die.h

@@ -11,7 +11,7 @@ inline void* memalign_or_die(size_t alignment, size_t size)
     void *buf = memalign(alignment, size);
     if (!buf)
     {
-        printf("Failed to allocate %lu bytes\n", size);
+        printf("Failed to allocate %zu bytes\n", size);
         exit(1);
     }
     return buf;
@@ -22,7 +22,7 @@ inline void* malloc_or_die(size_t size)
     void *buf = malloc(size);
     if (!buf)
     {
-        printf("Failed to allocate %lu bytes\n", size);
+        printf("Failed to allocate %zu bytes\n", size);
         exit(1);
     }
     return buf;
@@ -33,7 +33,7 @@ inline void* realloc_or_die(void *ptr, size_t size)
     void *buf = realloc(ptr, size);
     if (!buf)
     {
-        printf("Failed to allocate %lu bytes\n", size);
+        printf("Failed to allocate %zu bytes\n", size);
         exit(1);
     }
     return buf;
@@ -44,7 +44,7 @@ inline void* calloc_or_die(size_t nmemb, size_t size)
     void *buf = calloc(nmemb, size);
     if (!buf)
     {
-        printf("Failed to allocate %lu bytes\n", size * nmemb);
+        printf("Failed to allocate %zu bytes\n", size * nmemb);
         exit(1);
     }
     return buf;

src/messenger.cpp

@@ -27,13 +27,13 @@ void osd_messenger_t::init()
         if (!rdma_context)
         {
             if (log_level > 0)
-                fprintf(stderr, "[OSD %lu] Couldn't initialize RDMA, proceeding with TCP only\n", osd_num);
+                fprintf(stderr, "[OSD %ju] Couldn't initialize RDMA, proceeding with TCP only\n", osd_num);
         }
         else
         {
             rdma_max_sge = rdma_max_sge < rdma_context->attrx.orig_attr.max_sge
                 ? rdma_max_sge : rdma_context->attrx.orig_attr.max_sge;
-            fprintf(stderr, "[OSD %lu] RDMA initialized successfully\n", osd_num);
+            fprintf(stderr, "[OSD %ju] RDMA initialized successfully\n", osd_num);
             fcntl(rdma_context->channel->fd, F_SETFL, fcntl(rdma_context->channel->fd, F_GETFL, 0) | O_NONBLOCK);
             tfd->set_fd_handler(rdma_context->channel->fd, false, [this](int notify_fd, int epoll_events)
             {
@@ -45,11 +45,12 @@ void osd_messenger_t::init()
 #endif
     keepalive_timer_id = tfd->set_timer(1000, true, [this](int)
     {
-        std::vector<int> to_stop;
+        auto cl_it = clients.begin();
-        std::vector<osd_op_t*> to_ping;
+        while (cl_it != clients.end())
-        for (auto cl_it = clients.begin(); cl_it != clients.end(); cl_it++)
         {
             auto cl = cl_it->second;
+            cl_it++;
+            auto peer_fd = cl->peer_fd;
             if (!cl->osd_num || cl->peer_state != PEER_CONNECTED && cl->peer_state != PEER_RDMA)
             {
                 // Do not run keepalive on regular clients
@@ -61,8 +62,10 @@ void osd_messenger_t::init()
                 if (!cl->ping_time_remaining)
                 {
                     // Ping timed out, stop the client
-                    fprintf(stderr, "Ping timed out for OSD %lu (client %d), disconnecting peer\n", cl->osd_num, cl->peer_fd);
+                    fprintf(stderr, "Ping timed out for OSD %ju (client %d), disconnecting peer\n", cl->osd_num, cl->peer_fd);
-                    to_stop.push_back(cl->peer_fd);
+                    stop_client(peer_fd, true);
+                    // Restart iterator because it may be invalidated
+                    cl_it = clients.upper_bound(peer_fd);
                 }
             }
             else if (cl->idle_time_remaining > 0)
@@ -96,13 +99,15 @@ void osd_messenger_t::init()
                         delete op;
                         if (fail_fd >= 0)
                         {
-                            fprintf(stderr, "Ping failed for OSD %lu (client %d), disconnecting peer\n", fail_osd_num, fail_fd);
+                            fprintf(stderr, "Ping failed for OSD %ju (client %d), disconnecting peer\n", fail_osd_num, fail_fd);
                             stop_client(fail_fd, true);
                         }
                     };
-                    to_ping.push_back(op);
                     cl->ping_time_remaining = osd_ping_timeout;
                     cl->idle_time_remaining = osd_idle_timeout;
+                    outbox_push(op);
+                    // Restart iterator because it may be invalidated
+                    cl_it = clients.upper_bound(peer_fd);
                 }
             }
             else
@@ -110,15 +115,6 @@ void osd_messenger_t::init()
                 cl->idle_time_remaining = osd_idle_timeout;
             }
         }
-        // Don't stop clients while a 'clients' iterator is still active
-        for (int peer_fd: to_stop)
-        {
-            stop_client(peer_fd, true);
-        }
-        for (auto op: to_ping)
-        {
-            outbox_push(op);
-        }
     });
 }
 
@@ -257,7 +253,7 @@ void osd_messenger_t::try_connect_peer_addr(osd_num_t peer_osd, const char *peer
     clients[peer_fd] = new osd_client_t();
     if (log_level > 0)
     {
-        fprintf(stderr, "Connecting to OSD %lu at %s:%d (client %d)\n", peer_osd, peer_host, peer_port, peer_fd);
+        fprintf(stderr, "Connecting to OSD %ju at %s:%d (client %d)\n", peer_osd, peer_host, peer_port, peer_fd);
     }
     clients[peer_fd]->peer_addr = addr;
     clients[peer_fd]->peer_port = peer_port;
@@ -323,7 +319,7 @@ void osd_messenger_t::handle_peer_epoll(int peer_fd, int epoll_events)
         // Stop client
         if (log_level > 0)
         {
-            fprintf(stderr, "[OSD %lu] client %d disconnected\n", this->osd_num, peer_fd);
+            fprintf(stderr, "[OSD %ju] client %d disconnected\n", this->osd_num, peer_fd);
         }
         stop_client(peer_fd, true);
     }
@@ -349,7 +345,7 @@ void osd_messenger_t::on_connect_peer(osd_num_t peer_osd, int peer_fd)
     wp.connecting = false;
     if (peer_fd < 0)
     {
-        fprintf(stderr, "Failed to connect to peer OSD %lu address %s port %d: %s\n", peer_osd, wp.cur_addr.c_str(), wp.cur_port, strerror(-peer_fd));
+        fprintf(stderr, "Failed to connect to peer OSD %ju address %s port %d: %s\n", peer_osd, wp.cur_addr.c_str(), wp.cur_port, strerror(-peer_fd));
         if (wp.address_changed)
         {
             wp.address_changed = false;
@@ -376,7 +372,7 @@ void osd_messenger_t::on_connect_peer(osd_num_t peer_osd, int peer_fd)
     }
     if (log_level > 0)
     {
-        fprintf(stderr, "[OSD %lu] Connected with peer OSD %lu (client %d)\n", osd_num, peer_osd, peer_fd);
+        fprintf(stderr, "[OSD %ju] Connected with peer OSD %ju (client %d)\n", osd_num, peer_osd, peer_fd);
     }
     wanted_peers.erase(peer_osd);
     repeer_pgs(peer_osd);
@@ -422,7 +418,7 @@ void osd_messenger_t::check_peer_config(osd_client_t *cl)
         if (op->reply.hdr.retval < 0)
         {
             err = true;
-            fprintf(stderr, "Failed to get config from OSD %lu (retval=%ld), disconnecting peer\n", cl->osd_num, op->reply.hdr.retval);
+            fprintf(stderr, "Failed to get config from OSD %ju (retval=%jd), disconnecting peer\n", cl->osd_num, op->reply.hdr.retval);
         }
         else
         {
@@ -430,18 +426,18 @@ void osd_messenger_t::check_peer_config(osd_client_t *cl)
             if (json_err != "")
             {
                 err = true;
-                fprintf(stderr, "Failed to get config from OSD %lu: bad JSON: %s, disconnecting peer\n", cl->osd_num, json_err.c_str());
+                fprintf(stderr, "Failed to get config from OSD %ju: bad JSON: %s, disconnecting peer\n", cl->osd_num, json_err.c_str());
             }
             else if (config["osd_num"].uint64_value() != cl->osd_num)
             {
                 err = true;
-                fprintf(stderr, "Connected to OSD %lu instead of OSD %lu, peer state is outdated, disconnecting peer\n", config["osd_num"].uint64_value(), cl->osd_num);
+                fprintf(stderr, "Connected to OSD %ju instead of OSD %ju, peer state is outdated, disconnecting peer\n", config["osd_num"].uint64_value(), cl->osd_num);
             }
             else if (config["protocol_version"].uint64_value() != OSD_PROTOCOL_VERSION)
             {
                 err = true;
                 fprintf(
-                    stderr, "OSD %lu protocol version is %lu, but only version %u is supported.\n"
+                    stderr, "OSD %ju protocol version is %ju, but only version %u is supported.\n"
                     " If you need to upgrade from 0.5.x please request it via the issue tracker.\n",
                     cl->osd_num, config["protocol_version"].uint64_value(), OSD_PROTOCOL_VERSION
                 );
@@ -467,7 +463,7 @@ void osd_messenger_t::check_peer_config(osd_client_t *cl)
                 cl->rdma_conn->connect(&addr) != 0)
             {
                 fprintf(
-                    stderr, "Failed to connect to OSD %lu (address %s) using RDMA\n",
+                    stderr, "Failed to connect to OSD %ju (address %s) using RDMA\n",
                     cl->osd_num, config["rdma_address"].string_value().c_str()
                 );
                 delete cl->rdma_conn;
@@ -488,7 +484,7 @@ void osd_messenger_t::check_peer_config(osd_client_t *cl)
                 }
                 if (log_level > 0)
                 {
-                    fprintf(stderr, "Connected to OSD %lu using RDMA\n", cl->osd_num);
+                    fprintf(stderr, "Connected to OSD %ju using RDMA\n", cl->osd_num);
                 }
                 cl->peer_state = PEER_RDMA;
                 tfd->set_fd_handler(cl->peer_fd, false, [this](int peer_fd, int epoll_events)
@@ -520,7 +516,7 @@ void osd_messenger_t::accept_connections(int listen_fd)
     while ((peer_fd = accept(listen_fd, (sockaddr*)&addr, &peer_addr_size)) >= 0)
     {
         assert(peer_fd != 0);
-        fprintf(stderr, "[OSD %lu] new client %d: connection from %s\n", this->osd_num, peer_fd,
+        fprintf(stderr, "[OSD %ju] new client %d: connection from %s\n", this->osd_num, peer_fd,
             addr_to_string(addr).c_str());
         fcntl(peer_fd, F_SETFL, fcntl(peer_fd, F_GETFL, 0) | O_NONBLOCK);
         int one = 1;

src/msgr_op.h

@@ -76,7 +76,7 @@ struct osd_op_buf_list_t
                 buf = (iovec*)malloc(sizeof(iovec) * alloc);
                 if (!buf)
                 {
-                    fprintf(stderr, "Failed to allocate %lu bytes\n", sizeof(iovec) * alloc);
+                    fprintf(stderr, "Failed to allocate %u bytes\n", (int)sizeof(iovec) * alloc);
                     exit(1);
                 }
                 memcpy(buf, inline_buf, sizeof(iovec) * old);
@@ -87,7 +87,7 @@ struct osd_op_buf_list_t
                 buf = (iovec*)realloc(buf, sizeof(iovec) * alloc);
                 if (!buf)
                 {
-                    fprintf(stderr, "Failed to allocate %lu bytes\n", sizeof(iovec) * alloc);
+                    fprintf(stderr, "Failed to allocate %u bytes\n", (int)sizeof(iovec) * alloc);
                     exit(1);
                 }
             }
@@ -109,7 +109,7 @@ struct osd_op_buf_list_t
                 buf = (iovec*)malloc(sizeof(iovec) * alloc);
                 if (!buf)
                 {
-                    fprintf(stderr, "Failed to allocate %lu bytes\n", sizeof(iovec) * alloc);
+                    fprintf(stderr, "Failed to allocate %u bytes\n", (int)sizeof(iovec) * alloc);
                     exit(1);
                 }
                 memcpy(buf, inline_buf, sizeof(iovec)*old);
@@ -120,7 +120,7 @@ struct osd_op_buf_list_t
                 buf = (iovec*)realloc(buf, sizeof(iovec) * alloc);
                 if (!buf)
                 {
-                    fprintf(stderr, "Failed to allocate %lu bytes\n", sizeof(iovec) * alloc);
+                    fprintf(stderr, "Failed to allocate %u bytes\n", (int)sizeof(iovec) * alloc);
                     exit(1);
                 }
             }

src/msgr_rdma.cpp

@@ -10,7 +10,7 @@ std::string msgr_rdma_address_t::to_string()
 {
     char msg[sizeof "0000:00000000:00000000:00000000000000000000000000000000"];
     sprintf(
-        msg, "%04x:%06x:%06x:%016lx%016lx", lid, qpn, psn,
+        msg, "%04x:%06x:%06x:%016jx%016jx", lid, qpn, psn,
         htobe64(((uint64_t*)&gid)[0]), htobe64(((uint64_t*)&gid)[1])
     );
     return std::string(msg);
@@ -20,7 +20,7 @@ bool msgr_rdma_address_t::from_string(const char *str, msgr_rdma_address_t *dest
 {
     uint64_t* gid = (uint64_t*)&dest->gid;
     int scanned = sscanf(
-        str, "%hx:%x:%x:%16lx%16lx", &dest->lid, &dest->qpn, &dest->psn, gid, gid+1
+        str, "%hx:%x:%x:%16jx%16jx", &dest->lid, &dest->qpn, &dest->psn, gid, gid+1
     );
     gid[0] = be64toh(gid[0]);
     gid[1] = be64toh(gid[1]);
@@ -594,7 +594,7 @@ void osd_messenger_t::handle_rdma_events()
                 fprintf(stderr, "RDMA work request failed for client %d", client_id);
                 if (cl->osd_num)
                 {
-                    fprintf(stderr, " (OSD %lu)", cl->osd_num);
+                    fprintf(stderr, " (OSD %ju)", cl->osd_num);
                 }
                 fprintf(stderr, " with status: %s, stopping client\n", ibv_wc_status_str(wc[i].status));
                 stop_client(client_id);

src/msgr_receive.cpp

@@ -180,7 +180,7 @@ bool osd_messenger_t::handle_finished_read(osd_client_t *cl)
             handle_op_hdr(cl);
         else
         {
-            fprintf(stderr, "Received garbage: magic=%lx id=%lu opcode=%lx from %d\n", cl->read_op->req.hdr.magic, cl->read_op->req.hdr.id, cl->read_op->req.hdr.opcode, cl->peer_fd);
+            fprintf(stderr, "Received garbage: magic=%jx id=%ju opcode=%jx from %d\n", cl->read_op->req.hdr.magic, cl->read_op->req.hdr.id, cl->read_op->req.hdr.opcode, cl->peer_fd);
             stop_client(cl->peer_fd);
             return false;
         }
@@ -297,7 +297,7 @@ bool osd_messenger_t::handle_reply_hdr(osd_client_t *cl)
     if (req_it == cl->sent_ops.end())
     {
         // Command out of sync. Drop connection
-        fprintf(stderr, "Client %d command out of sync: id %lu\n", cl->peer_fd, cl->read_op->req.hdr.id);
+        fprintf(stderr, "Client %d command out of sync: id %ju\n", cl->peer_fd, cl->read_op->req.hdr.id);
         stop_client(cl->peer_fd);
         return false;
     }
@@ -312,7 +312,7 @@ bool osd_messenger_t::handle_reply_hdr(osd_client_t *cl)
         if (op->reply.hdr.retval >= 0 && (op->reply.hdr.retval != expected_size || bmp_len > op->bitmap_len))
         {
             // Check reply length to not overflow the buffer
-            fprintf(stderr, "Client %d read reply of different length: expected %u+%u, got %ld+%u\n",
+            fprintf(stderr, "Client %d read reply of different length: expected %u+%u, got %jd+%u\n",
                 cl->peer_fd, expected_size, op->bitmap_len, op->reply.hdr.retval, bmp_len);
             cl->sent_ops[op->req.hdr.id] = op;
             stop_client(cl->peer_fd);

src/msgr_stop.cpp

@@ -61,11 +61,11 @@ void osd_messenger_t::stop_client(int peer_fd, bool force, bool force_delete)
     {
         if (cl->osd_num)
         {
-            fprintf(stderr, "[OSD %lu] Stopping client %d (OSD peer %lu)\n", osd_num, peer_fd, cl->osd_num);
+            fprintf(stderr, "[OSD %ju] Stopping client %d (OSD peer %ju)\n", osd_num, peer_fd, cl->osd_num);
         }
         else
         {
-            fprintf(stderr, "[OSD %lu] Stopping client %d (regular client)\n", osd_num, peer_fd);
+            fprintf(stderr, "[OSD %ju] Stopping client %d (regular client)\n", osd_num, peer_fd);
         }
     }
     // First set state to STOPPED so another stop_client() call doesn't try to free it again

src/nbd_proxy.cpp

@@ -738,7 +738,7 @@ protected:
             }
             uint64_t handle = *((uint64_t*)cur_req.handle);
 #ifdef DEBUG
-            printf("request %lx +%x %lx\n", be64toh(cur_req.from), be32toh(cur_req.len), handle);
+            printf("request %jx +%x %jx\n", be64toh(cur_req.from), be32toh(cur_req.len), handle);
 #endif
             void *buf = NULL;
             cluster_op_t *op = new cluster_op_t;
@@ -759,7 +759,7 @@ protected:
             op->callback = [this, buf, handle](cluster_op_t *op)
             {
 #ifdef DEBUG
-                printf("reply %lx e=%d\n", handle, op->retval);
+                printf("reply %jx e=%d\n", handle, op->retval);
 #endif
                 nbd_reply *reply = (nbd_reply*)buf;
                 reply->magic = htobe32(NBD_REPLY_MAGIC);
@@ -769,7 +769,7 @@ protected:
                 if (op->retval < 0 || op->opcode != OSD_OP_READ)
                     to_list.push_back({ .iov_base = buf, .iov_len = sizeof(nbd_reply) });
                 else
-                    to_list.push_back({ .iov_base = buf, .iov_len = sizeof(nbd_reply) + op->len });
+                    to_list.push_back({ .iov_base = buf, .iov_len = sizeof(nbd_reply) + (size_t)op->len });
                 to_free.push_back(buf);
                 delete op;
                 ringloop->wakeup();

src/nfs_conn.cpp

@@ -517,7 +517,7 @@ static void extend_inode(nfs_client_t *self, uint64_t inode, uint64_t new_size)
             auto & ext = self->extends[inode];
             if (r.err)
             {
-                fprintf(stderr, "Error extending inode %lu to %lu bytes: %s\n", inode, new_size, r.text.c_str());
+                fprintf(stderr, "Error extending inode %ju to %ju bytes: %s\n", inode, new_size, r.text.c_str());
             }
             if (r.err == EAGAIN || ext.next_extend > ext.cur_extend)
             {

src/nfs_proxy.cpp

@@ -350,7 +350,7 @@ void nfs_proxy_t::parse_stats(etcd_kv_t & kv)
         pool_id_t pool_id = 0;
         inode_t inode_num = 0;
         char null_byte = 0;
-        int scanned = sscanf(key.c_str() + cli->st_cli.etcd_prefix.length()+13, "%u/%lu%c", &pool_id, &inode_num, &null_byte);
+        int scanned = sscanf(key.c_str() + cli->st_cli.etcd_prefix.length()+13, "%u/%ju%c", &pool_id, &inode_num, &null_byte);
         if (scanned != 2 || !pool_id || pool_id >= POOL_ID_MAX || !inode_num)
         {
             fprintf(stderr, "Bad etcd key %s, ignoring\n", key.c_str());
@@ -387,7 +387,7 @@ void nfs_proxy_t::check_default_pool()
         }
         else
         {
-            fprintf(stderr, "There are %lu pools. Please select default pool with --pool option\n", cli->st_cli.pool_config.size());
+            fprintf(stderr, "There are %zu pools. Please select default pool with --pool option\n", cli->st_cli.pool_config.size());
             exit(1);
         }
     }

src/osd.cpp

@@ -48,6 +48,15 @@ osd_t::osd_t(const json11::Json & config, ring_loop_t *ringloop)
     {
         auto bs_cfg = json_to_bs(this->config);
         this->bs = new blockstore_t(bs_cfg, ringloop, tfd);
+        // Wait for blockstore initialisation before actually starting OSD logic
+        // to prevent peering timeouts during restart with filled databases
+        while (!bs->is_started())
+        {
+            ringloop->loop();
+            if (bs->is_started())
+                break;
+            ringloop->wait();
+        }
         // Autosync based on the number of unstable writes to prevent stalls due to insufficient journal space
         uint64_t max_autosync = bs->get_journal_size() / bs->get_block_size() / 2;
         if (autosync_writes > max_autosync)
@@ -233,6 +242,8 @@ void osd_t::parse_config(bool init)
         ? 10 : config["recovery_tune_agg_interval"].uint64_value();
     recovery_tune_sleep_min_us = config["recovery_tune_sleep_min_us"].is_null()
         ? 10 : config["recovery_tune_sleep_min_us"].uint64_value();
+    recovery_tune_sleep_cutoff_us = config["recovery_tune_sleep_cutoff_us"].is_null()
+        ? 10000000 : config["recovery_tune_sleep_cutoff_us"].uint64_value();
     recovery_pg_switch = config["recovery_pg_switch"].uint64_value();
     if (recovery_pg_switch < 1)
         recovery_pg_switch = DEFAULT_RECOVERY_PG_SWITCH;
@@ -473,14 +484,14 @@ void osd_t::print_stats()
             if (msgr.stats.op_stat_bytes[i] != 0)
             {
                 printf(
-                    "[OSD %lu] avg latency for op %d (%s): %lu us, B/W: %.2f %s\n", osd_num, i, osd_op_names[i], avg,
+                    "[OSD %ju] avg latency for op %d (%s): %ju us, B/W: %.2f %s\n", osd_num, i, osd_op_names[i], avg,
                     (bw > 1024*1024*1024 ? bw/1024.0/1024/1024 : (bw > 1024*1024 ? bw/1024.0/1024 : bw/1024.0)),
                     (bw > 1024*1024*1024 ? "GB/s" : (bw > 1024*1024 ? "MB/s" : "KB/s"))
                 );
             }
             else
             {
-                printf("[OSD %lu] avg latency for op %d (%s): %lu us\n", osd_num, i, osd_op_names[i], avg);
+                printf("[OSD %ju] avg latency for op %d (%s): %ju us\n", osd_num, i, osd_op_names[i], avg);
             }
             prev_stats.op_stat_count[i] = msgr.stats.op_stat_count[i];
             prev_stats.op_stat_sum[i] = msgr.stats.op_stat_sum[i];
@@ -492,7 +503,7 @@ void osd_t::print_stats()
         if (msgr.stats.subop_stat_count[i] != prev_stats.subop_stat_count[i])
         {
             uint64_t avg = (msgr.stats.subop_stat_sum[i] - prev_stats.subop_stat_sum[i])/(msgr.stats.subop_stat_count[i] - prev_stats.subop_stat_count[i]);
-            printf("[OSD %lu] avg latency for subop %d (%s): %ld us\n", osd_num, i, osd_op_names[i], avg);
+            printf("[OSD %ju] avg latency for subop %d (%s): %jd us\n", osd_num, i, osd_op_names[i], avg);
             prev_stats.subop_stat_count[i] = msgr.stats.subop_stat_count[i];
             prev_stats.subop_stat_sum[i] = msgr.stats.subop_stat_sum[i];
         }
@@ -503,7 +514,7 @@ void osd_t::print_stats()
         {
             uint64_t bw = (recovery_stat[i].bytes - recovery_print_prev[i].bytes) / print_stats_interval;
             printf(
-                "[OSD %lu] %s recovery: %.1f op/s, B/W: %.2f %s, avg latency %ld us, delay %ld us\n", osd_num, recovery_stat_names[i],
+                "[OSD %ju] %s recovery: %.1f op/s, B/W: %.2f %s, avg latency %jd us, delay %jd us\n", osd_num, recovery_stat_names[i],
                 (recovery_stat[i].count - recovery_print_prev[i].count) * 1.0 / print_stats_interval,
                 (bw > 1024*1024*1024 ? bw/1024.0/1024/1024 : (bw > 1024*1024 ? bw/1024.0/1024 : bw/1024.0)),
                 (bw > 1024*1024*1024 ? "GB/s" : (bw > 1024*1024 ? "MB/s" : "KB/s")),
@@ -515,19 +526,19 @@ void osd_t::print_stats()
     memcpy(recovery_print_prev, recovery_stat, sizeof(recovery_stat));
     if (corrupted_objects > 0)
     {
-        printf("[OSD %lu] %lu object(s) corrupted\n", osd_num, corrupted_objects);
+        printf("[OSD %ju] %ju object(s) corrupted\n", osd_num, corrupted_objects);
     }
     if (incomplete_objects > 0)
     {
-        printf("[OSD %lu] %lu object(s) incomplete\n", osd_num, incomplete_objects);
+        printf("[OSD %ju] %ju object(s) incomplete\n", osd_num, incomplete_objects);
     }
     if (degraded_objects > 0)
     {
-        printf("[OSD %lu] %lu object(s) degraded\n", osd_num, degraded_objects);
+        printf("[OSD %ju] %ju object(s) degraded\n", osd_num, degraded_objects);
     }
     if (misplaced_objects > 0)
     {
-        printf("[OSD %lu] %lu object(s) misplaced\n", osd_num, misplaced_objects);
+        printf("[OSD %ju] %ju object(s) misplaced\n", osd_num, misplaced_objects);
     }
 }
 
@@ -546,27 +557,27 @@ void osd_t::print_slow()
                 int l = sizeof(alloc), n;
                 char *buf = alloc;
 #define bufprintf(s, ...) { n = snprintf(buf, l, s, __VA_ARGS__); n = n < 0 ? 0 : n; buf += n; l -= n; }
-                bufprintf("[OSD %lu] Slow op %lx", osd_num, (unsigned long)op);
+                bufprintf("[OSD %ju] Slow op %jx", osd_num, (uint64_t)op);
                 if (kv.second->osd_num)
                 {
-                    bufprintf(" from peer OSD %lu (client %d)", kv.second->osd_num, kv.second->peer_fd);
+                    bufprintf(" from peer OSD %ju (client %d)", kv.second->osd_num, kv.second->peer_fd);
                 }
                 else
                 {
                     bufprintf(" from client %d", kv.second->peer_fd);
                 }
-                bufprintf(": %s id=%lu", osd_op_names[op->req.hdr.opcode], op->req.hdr.id);
+                bufprintf(": %s id=%ju", osd_op_names[op->req.hdr.opcode], op->req.hdr.id);
                 if (op->req.hdr.opcode == OSD_OP_SEC_READ || op->req.hdr.opcode == OSD_OP_SEC_WRITE ||
                     op->req.hdr.opcode == OSD_OP_SEC_WRITE_STABLE || op->req.hdr.opcode == OSD_OP_SEC_DELETE)
                 {
-                    bufprintf(" %lx:%lx v", op->req.sec_rw.oid.inode, op->req.sec_rw.oid.stripe);
+                    bufprintf(" %jx:%jx v", op->req.sec_rw.oid.inode, op->req.sec_rw.oid.stripe);
                     if (op->req.sec_rw.version == UINT64_MAX)
                     {
                         bufprintf("%s", "max");
                     }
                     else
                     {
-                        bufprintf("%lu", op->req.sec_rw.version);
+                        bufprintf("%ju", op->req.sec_rw.version);
                     }
                     if (op->req.hdr.opcode != OSD_OP_SEC_DELETE)
                     {
@@ -578,17 +589,17 @@ void osd_t::print_slow()
                     for (uint64_t i = 0; i < op->req.sec_stab.len && i < sizeof(obj_ver_id)*12; i += sizeof(obj_ver_id))
                     {
                         obj_ver_id *ov = (obj_ver_id*)((uint8_t*)op->buf + i);
-                        bufprintf(i == 0 ? " %lx:%lx v%lu" : ", %lx:%lx v%lu", ov->oid.inode, ov->oid.stripe, ov->version);
+                        bufprintf(i == 0 ? " %jx:%jx v%ju" : ", %jx:%jx v%ju", ov->oid.inode, ov->oid.stripe, ov->version);
                     }
                     if (op->req.sec_stab.len > sizeof(obj_ver_id)*12)
                     {
-                        bufprintf(", ... (%lu items)", op->req.sec_stab.len/sizeof(obj_ver_id));
+                        bufprintf(", ... (%ju items)", op->req.sec_stab.len/sizeof(obj_ver_id));
                     }
                 }
                 else if (op->req.hdr.opcode == OSD_OP_SEC_LIST)
                 {
                     bufprintf(
-                        " oid=%lx/%lx-%lx/%lx pg=%u/%u, stripe=%lu, limit=%u",
+                        " oid=%jx/%jx-%jx/%jx pg=%u/%u, stripe=%ju, limit=%u",
                         op->req.sec_list.min_inode, op->req.sec_list.min_stripe,
                         op->req.sec_list.max_inode, op->req.sec_list.max_stripe,
                         op->req.sec_list.list_pg, op->req.sec_list.pg_count,
@@ -598,7 +609,7 @@ void osd_t::print_slow()
                 else if (op->req.hdr.opcode == OSD_OP_READ || op->req.hdr.opcode == OSD_OP_WRITE ||
                     op->req.hdr.opcode == OSD_OP_DELETE)
                 {
-                    bufprintf(" inode=%lx offset=%lx len=%x", op->req.rw.inode, op->req.rw.offset, op->req.rw.len);
+                    bufprintf(" inode=%jx offset=%jx len=%x", op->req.rw.inode, op->req.rw.offset, op->req.rw.len);
                 }
                 if (op->req.hdr.opcode == OSD_OP_SEC_READ || op->req.hdr.opcode == OSD_OP_SEC_WRITE ||
                     op->req.hdr.opcode == OSD_OP_SEC_WRITE_STABLE || op->req.hdr.opcode == OSD_OP_SEC_DELETE ||
@@ -610,7 +621,7 @@ void osd_t::print_slow()
                     int wait_for = op->bs_op ? PRIV(op->bs_op)->wait_for : 0;
                     if (wait_for)
                     {
-                        bufprintf(" wait=%d (detail=%lu)", wait_for, PRIV(op->bs_op)->wait_detail);
+                        bufprintf(" wait=%d (detail=%ju)", wait_for, PRIV(op->bs_op)->wait_detail);
                     }
                 }
                 else if (op->req.hdr.opcode == OSD_OP_READ || op->req.hdr.opcode == OSD_OP_WRITE ||

src/osd.h

@@ -125,6 +125,7 @@ class osd_t
     int recovery_tune_interval = 1;
     int recovery_tune_agg_interval = 10;
     int recovery_tune_sleep_min_us = 10;
+    int recovery_tune_sleep_cutoff_us = 10000000;
     int recovery_pg_switch = DEFAULT_RECOVERY_PG_SWITCH;
     int recovery_sync_batch = DEFAULT_RECOVERY_BATCH;
     int inode_vanish_time = 60;
@@ -282,6 +283,7 @@ class osd_t
     void exec_sync_stab_all(osd_op_t *cur_op);
     void exec_show_config(osd_op_t *cur_op);
     void exec_secondary(osd_op_t *cur_op);
+    void exec_secondary_real(osd_op_t *cur_op);
     void secondary_op_callback(osd_op_t *cur_op);
 
     // primary ops

src/osd_cluster.cpp

@@ -117,7 +117,7 @@ bool osd_t::check_peer_config(osd_client_t *cl, json11::Json conf)
         conf["immediate_commit"].is_null())
     {
         printf(
-            "[OSD %lu] Warning: peer OSD %lu does not report block_size/bitmap_granularity/immediate_commit."
+            "[OSD %ju] Warning: peer OSD %ju does not report block_size/bitmap_granularity/immediate_commit."
             " Is it older than 0.6.3?\n", this->osd_num, cl->osd_num
         );
     }
@@ -129,7 +129,7 @@ bool osd_t::check_peer_config(osd_client_t *cl, json11::Json conf)
             immediate_commit == IMMEDIATE_SMALL && peer_immediate_commit == IMMEDIATE_NONE)
         {
             printf(
-                "[OSD %lu] My immediate_commit is \"%s\", but peer OSD %lu has \"%s\". We can't work together\n",
+                "[OSD %ju] My immediate_commit is \"%s\", but peer OSD %ju has \"%s\". We can't work together\n",
                 this->osd_num, immediate_commit == IMMEDIATE_ALL ? "all" : "small",
                 cl->osd_num, conf["immediate_commit"].string_value().c_str()
             );
@@ -138,7 +138,7 @@ bool osd_t::check_peer_config(osd_client_t *cl, json11::Json conf)
         else if (conf["block_size"].uint64_value() != (uint64_t)this->bs_block_size)
         {
             printf(
-                "[OSD %lu] My block_size is %u, but peer OSD %lu has %lu. We can't work together\n",
+                "[OSD %ju] My block_size is %u, but peer OSD %ju has %ju. We can't work together\n",
                 this->osd_num, this->bs_block_size, cl->osd_num, conf["block_size"].uint64_value()
             );
             return false;
@@ -146,7 +146,7 @@ bool osd_t::check_peer_config(osd_client_t *cl, json11::Json conf)
         else if (conf["bitmap_granularity"].uint64_value() != (uint64_t)this->bs_bitmap_granularity)
         {
             printf(
-                "[OSD %lu] My bitmap_granularity is %u, but peer OSD %lu has %lu. We can't work together\n",
+                "[OSD %ju] My bitmap_granularity is %u, but peer OSD %ju has %ju. We can't work together\n",
                 this->osd_num, this->bs_bitmap_granularity, cl->osd_num, conf["bitmap_granularity"].uint64_value()
             );
             return false;
@@ -181,7 +181,7 @@ json11::Json osd_t::get_statistics()
     timespec ts;
     clock_gettime(CLOCK_REALTIME, &ts);
     char time_str[50] = { 0 };
-    sprintf(time_str, "%ld.%03ld", ts.tv_sec, ts.tv_nsec/1000000);
+    sprintf(time_str, "%jd.%03ld", (uint64_t)ts.tv_sec, ts.tv_nsec/1000000);
     st["time"] = time_str;
     if (bs)
     {
@@ -358,7 +358,7 @@ void osd_t::report_statistics()
         etcd_reporting_stats = false;
         if (err != "")
         {
-            printf("[OSD %lu] Error reporting state to etcd: %s\n", this->osd_num, err.c_str());
+            printf("[OSD %ju] Error reporting state to etcd: %s\n", this->osd_num, err.c_str());
             // Retry indefinitely
             tfd->set_timer(st_cli.etcd_slow_timeout, false, [this](int timer_id)
             {
@@ -367,7 +367,7 @@ void osd_t::report_statistics()
         }
         else if (res["error"].string_value() != "")
         {
-            printf("[OSD %lu] Error reporting state to etcd: %s\n", this->osd_num, res["error"].string_value().c_str());
+            printf("[OSD %ju] Error reporting state to etcd: %s\n", this->osd_num, res["error"].string_value().c_str());
             force_stop(1);
         }
     });
@@ -432,7 +432,7 @@ void osd_t::acquire_lease()
         create_osd_state();
     });
     printf(
-        "[OSD %lu] reporting to etcd at %s every %d seconds (statistics every %d seconds)\n", this->osd_num,
+        "[OSD %ju] reporting to etcd at %s every %d seconds (statistics every %d seconds)\n", this->osd_num,
         (config["etcd_address"].is_string() ? config["etcd_address"].string_value() : config["etcd_address"].dump()).c_str(),
         etcd_report_interval, etcd_stats_interval
     );
@@ -499,11 +499,11 @@ void osd_t::create_osd_state()
         {
             // OSD is already up
             auto kv = st_cli.parse_etcd_kv(data["responses"][0]["response_range"]["kvs"][0]);
-            printf("Key %s already exists in etcd, OSD %lu is still up\n", kv.key.c_str(), this->osd_num);
+            printf("Key %s already exists in etcd, OSD %ju is still up\n", kv.key.c_str(), this->osd_num);
             int64_t port = kv.value["port"].int64_value();
             for (auto & addr: kv.value["addresses"].array_items())
             {
-                printf("  listening at: %s:%ld\n", addr.string_value().c_str(), port);
+                printf("  listening at: %s:%jd\n", addr.string_value().c_str(), port);
             }
             force_stop(0);
             return;
@@ -569,13 +569,13 @@ void osd_t::force_stop(int exitcode)
             {
                 printf("Error revoking etcd lease: %s\n", err.c_str());
             }
-            printf("[OSD %lu] Force stopping\n", this->osd_num);
+            printf("[OSD %ju] Force stopping\n", this->osd_num);
             exit(exitcode);
         });
     }
     else
     {
-        printf("[OSD %lu] Force stopping\n", this->osd_num);
+        printf("[OSD %ju] Force stopping\n", this->osd_num);
         exit(exitcode);
     }
 }
@@ -629,7 +629,7 @@ void osd_t::apply_pg_count()
             if (still_active > 0)
             {
                 printf(
-                    "[OSD %lu] PG count change detected for pool %u (new is %lu, old is %u),"
+                    "[OSD %ju] PG count change detected for pool %u (new is %ju, old is %u),"
                     " but %u PG(s) are still active. This is not allowed. Exiting\n",
                     this->osd_num, pool_item.first, pool_item.second.real_pg_count, pg_counts[pool_item.first], still_active
                 );
@@ -663,7 +663,7 @@ void osd_t::apply_pg_config()
                 if (!warned_block_size)
                 {
                     printf(
-                        "[OSD %lu] My block_size and bitmap_granularity are %u/%u"
+                        "[OSD %ju] My block_size and bitmap_granularity are %u/%u"
                         ", but pool %u has %u/%u. Refusing to start PGs of this pool\n",
                         this->osd_num, bs_block_size, bs_bitmap_granularity,
                         pool_id, pool_item.second.data_block_size, pool_item.second.bitmap_granularity
@@ -843,7 +843,13 @@ void osd_t::report_pg_states()
                     pg_state_exists = true;
                     if (pg.state == PG_OFFLINE && pg_it->second.cur_primary != this->osd_num)
                     {
-                        // Nothing to check or report, PG is already taken over by another OSD
+                        // Nothing to report, PG is already taken over by another OSD
+                        checks.push_back(json11::Json::object {
+                            { "target", "MOD" },
+                            { "key", state_key_base64 },
+                            { "result", "LESS" },
+                            { "mod_revision", st_cli.etcd_watch_revision+1 },
+                        });
                         continue;
                     }
                 }
@@ -851,11 +857,6 @@ void osd_t::report_pg_states()
         }
         if (!pg_state_exists)
         {
-            if (pg.state == PG_OFFLINE)
-            {
-                // Nothing to check or report, PG is already stopped
-                continue;
-            }
             // Check that the PG key does not exist
             // Failed check indicates an unsuccessful PG lock attempt in this case
             checks.push_back(json11::Json::object {
@@ -984,7 +985,7 @@ void osd_t::report_pg_states()
                                 kv.value["primary"].uint64_value() != this->osd_num)
                             {
                                 // PG is somehow captured by another OSD
-                                printf("BUG: OSD %lu captured our PG %u/%u. Race condition detected, exiting\n",
+                                printf("BUG: OSD %ju captured our PG %u/%u. Race condition detected, exiting\n",
                                     kv.value["primary"].uint64_value(), pool_id, pg_num);
                                 force_stop(1);
                                 return;

src/osd_flush.cpp

@@ -66,7 +66,7 @@ void osd_t::handle_flush_op(bool rollback, pool_id_t pool_id, pg_num_t pg_num, p
 {
     if (log_level > 2)
     {
-        printf("[PG %u/%u] flush batch %lx completed on OSD %lu with result %d\n",
+        printf("[PG %u/%u] flush batch %jx completed on OSD %ju with result %d\n",
             pool_id, pg_num, (uint64_t)fb, peer_osd, retval);
     }
     pool_pg_num_t pg_id = { .pool_id = pool_id, .pg_num = pg_num };
@@ -88,7 +88,7 @@ void osd_t::handle_flush_op(bool rollback, pool_id_t pool_id, pg_num_t pg_num, p
         }
         else
         {
-            printf("Error while doing flush on OSD %lu: %d (%s)\n", osd_num, retval, strerror(-retval));
+            printf("Error while doing flush on OSD %ju: %d (%s)\n", osd_num, retval, strerror(-retval));
             auto fd_it = msgr.osd_peer_fds.find(peer_osd);
             if (fd_it != msgr.osd_peer_fds.end())
             {
@@ -122,7 +122,7 @@ void osd_t::handle_flush_op(bool rollback, pool_id_t pool_id, pg_num_t pg_num, p
                 {
                     if (log_level > 2)
                     {
-                        printf("[PG %u/%u] continuing write %lx to object %lx:%lx after flush\n",
+                        printf("[PG %u/%u] continuing write %jx to object %jx:%jx after flush\n",
                             pool_id, pg_num, (uint64_t)wr_it->second, wr_it->first.inode, wr_it->first.stripe);
                     }
                     continue_ops.push_back(wr_it->second);
@@ -169,12 +169,12 @@ bool osd_t::submit_flush_op(pool_id_t pool_id, pg_num_t pg_num, pg_flush_batch_t
     if (log_level > 2)
     {
         printf(
-            "[PG %u/%u] flush batch %lx on OSD %lu: %s objects: ",
+            "[PG %u/%u] flush batch %jx on OSD %ju: %s objects: ",
             pool_id, pg_num, (uint64_t)fb, peer_osd, rollback ? "rollback" : "stabilize"
         );
         for (int i = 0; i < count; i++)
         {
-            printf(i > 0 ? ", %lx:%lx v%lu" : "%lx:%lx v%lu", data[i].oid.inode, data[i].oid.stripe, data[i].version);
+            printf(i > 0 ? ", %jx:%jx v%ju" : "%jx:%jx v%ju", data[i].oid.inode, data[i].oid.stripe, data[i].version);
         }
         printf("\n");
     }
@@ -305,7 +305,7 @@ void osd_t::submit_recovery_op(osd_recovery_op_t *op)
     };
     if (log_level > 2)
     {
-        printf("Submitting recovery operation for %lx:%lx (%s)\n", op->oid.inode, op->oid.stripe, op->degraded ? "degraded" : "misplaced");
+        printf("Submitting recovery operation for %jx:%jx (%s)\n", op->oid.inode, op->oid.stripe, op->degraded ? "degraded" : "misplaced");
     }
     op->osd_op->peer_fd = -1;
     op->osd_op->callback = [this, op](osd_op_t *osd_op)
@@ -315,7 +315,7 @@ void osd_t::submit_recovery_op(osd_recovery_op_t *op)
             // Error recovering object
             // EPIPE is totally harmless (peer is gone), others like EIO/EDOM may be not
             printf(
-                "[PG %u/%u] Recovery operation failed with object %lx:%lx: error %ld\n",
+                "[PG %u/%u] Recovery operation failed with object %jx:%jx: error %jd\n",
                 INODE_POOL(op->oid.inode),
                 map_to_pg(op->oid, st_cli.pool_config.at(INODE_POOL(op->oid.inode)).pg_stripe_size),
                 op->oid.inode, op->oid.stripe, osd_op->reply.hdr.retval
@@ -323,7 +323,7 @@ void osd_t::submit_recovery_op(osd_recovery_op_t *op)
         }
         else if (log_level > 2)
         {
-            printf("Recovery operation done for %lx:%lx\n", op->oid.inode, op->oid.stripe);
+            printf("Recovery operation done for %jx:%jx\n", op->oid.inode, op->oid.stripe);
         }
         finish_recovery_op(op);
     };
@@ -422,6 +422,10 @@ void osd_t::tune_recovery()
     rtune_avg_lat = total_recovery_usec/recovery_count;
     uint64_t target_lat = rtune_avg_lat * rtune_avg_lat/1000000.0 * recovery_count/recovery_tune_interval / rtune_target_util;
     auto sleep_us = target_lat > rtune_avg_lat+recovery_tune_sleep_min_us ? target_lat-rtune_avg_lat : 0;
+    if (sleep_us > recovery_tune_sleep_cutoff_us)
+    {
+        return;
+    }
     if (recovery_target_sleep_items.size() != recovery_tune_agg_interval)
     {
         recovery_target_sleep_items.resize(recovery_tune_agg_interval);
@@ -438,10 +442,10 @@ void osd_t::tune_recovery()
     if (recovery_target_sleep_count < recovery_tune_agg_interval)
         recovery_target_sleep_count++;
     recovery_target_sleep_us = recovery_target_sleep_total / recovery_target_sleep_count;
-    if (log_level > 4)
+    if (log_level > 1)
     {
         printf(
-            "[OSD %lu] auto-tune: client util: %.2f, recovery util: %.2f, lat: %lu us -> target util %.2f, delay %lu us\n",
+            "[OSD %ju] auto-tune: client util: %.2f, recovery util: %.2f, lat: %ju us -> target util %.2f, delay %ju us\n",
             osd_num, rtune_client_util, total_recovery_usec/1000000.0/recovery_tune_interval,
             rtune_avg_lat, rtune_target_util, recovery_target_sleep_us
         );

src/osd_peering.cpp

@@ -113,7 +113,7 @@ void osd_t::repeer_pgs(osd_num_t peer_osd)
             if (repeer)
             {
                 // Repeer this pg
-                printf("[PG %u/%u] Repeer because of OSD %lu\n", pg.pool_id, pg.pg_num, peer_osd);
+                printf("[PG %u/%u] Repeer because of OSD %ju\n", pg.pool_id, pg.pg_num, peer_osd);
                 if (!(pg.state & (PG_ACTIVE | PG_REPEERING)) || pg.inflight == 0 && !pg.flush_batch)
                 {
                     start_pg_peering(pg);
@@ -222,6 +222,9 @@ void osd_t::start_pg_peering(pg_t & pg)
     }
     if (pg.pg_cursize < pg.pg_minsize)
     {
+        // FIXME: Incomplete EC PGs may currently easily lead to write hangs ("slow ops" in OSD logs)
+        // because such PGs don't flush unstable entries on secondary OSDs so they can't remove these
+        // entries from their journals...
         pg.state = PG_INCOMPLETE;
         report_pg_state(pg);
         return;
@@ -344,7 +347,7 @@ void osd_t::submit_list_subop(osd_num_t role_osd, pg_peering_state_t *ps)
             }
             add_bs_subop_stats(op);
             printf(
-                "[PG %u/%u] Got object list from OSD %lu (local): %d object versions (%lu of them stable)\n",
+                "[PG %u/%u] Got object list from OSD %ju (local): %d object versions (%ju of them stable)\n",
                 ps->pool_id, ps->pg_num, role_osd, bs_op->retval, bs_op->version
             );
             ps->list_results[role_osd] = {
@@ -384,7 +387,7 @@ void osd_t::submit_list_subop(osd_num_t role_osd, pg_peering_state_t *ps)
         {
             if (op->reply.hdr.retval < 0)
             {
-                printf("Failed to get object list from OSD %lu (retval=%ld), disconnecting peer\n", role_osd, op->reply.hdr.retval);
+                printf("Failed to get object list from OSD %ju (retval=%jd), disconnecting peer\n", role_osd, op->reply.hdr.retval);
                 int fail_fd = op->peer_fd;
                 ps->list_ops.erase(role_osd);
                 delete op;
@@ -392,7 +395,7 @@ void osd_t::submit_list_subop(osd_num_t role_osd, pg_peering_state_t *ps)
                 return;
             }
             printf(
-                "[PG %u/%u] Got object list from OSD %lu: %ld object versions (%lu of them stable)\n",
+                "[PG %u/%u] Got object list from OSD %ju: %jd object versions (%ju of them stable)\n",
                 ps->pool_id, ps->pg_num, role_osd, op->reply.hdr.retval, op->reply.sec_list.stable_count
             );
             ps->list_results[role_osd] = {

src/osd_peering_pg.cpp

@@ -239,7 +239,7 @@ void pg_obj_state_check_t::finish_object()
     {
         if (log_level > 1)
         {
-            printf("Object is incomplete: %lx:%lx version=%lu/%lu\n", oid.inode, oid.stripe, target_ver, max_ver);
+            printf("Object is incomplete: %jx:%jx version=%ju/%ju\n", oid.inode, oid.stripe, target_ver, max_ver);
         }
         state = OBJ_INCOMPLETE;
         pg->state = pg->state | PG_HAS_INCOMPLETE;
@@ -248,7 +248,7 @@ void pg_obj_state_check_t::finish_object()
     {
         if (log_level > 1)
         {
-            printf("Object is degraded: %lx:%lx version=%lu/%lu\n", oid.inode, oid.stripe, target_ver, max_ver);
+            printf("Object is degraded: %jx:%jx version=%ju/%ju\n", oid.inode, oid.stripe, target_ver, max_ver);
         }
         state = OBJ_DEGRADED;
         pg->state = pg->state | PG_HAS_DEGRADED;
@@ -257,7 +257,7 @@ void pg_obj_state_check_t::finish_object()
     {
         if (log_level > 2)
         {
-            printf("Object is misplaced: %lx:%lx version=%lu/%lu\n", oid.inode, oid.stripe, target_ver, max_ver);
+            printf("Object is misplaced: %jx:%jx version=%ju/%ju\n", oid.inode, oid.stripe, target_ver, max_ver);
         }
         state |= OBJ_MISPLACED;
         pg->state = pg->state | PG_HAS_MISPLACED;
@@ -267,7 +267,7 @@ void pg_obj_state_check_t::finish_object()
     {
         for (int i = obj_start; i < obj_end; i++)
         {
-            printf("v%lu present on: osd %lu, role %ld%s\n", list[i].version, list[i].osd_num,
+            printf("v%ju present on: osd %ju, role %jd%s\n", list[i].version, list[i].osd_num,
                 (list[i].oid.stripe & STRIPE_MASK), list[i].is_stable ? " (stable)" : "");
         }
     }
@@ -445,7 +445,7 @@ void pg_t::calc_object_states(int log_level)
             osd_set_desc += (osd_set_desc == "" ? "" : ", ")+std::to_string(osd_num);
         }
         printf(
-            "[PG %u/%u] %lu clean objects on target OSD set %s\n",
+            "[PG %u/%u] %ju clean objects on target OSD set %s\n",
             pool_id, pg_num, clean_count, osd_set_desc.c_str()
         );
         for (auto & stp: state_dict)
@@ -460,7 +460,7 @@ void pg_t::calc_object_states(int log_level)
                     (loc.loc_bad & LOC_CORRUPTED ? "(corrupted)" : "")+
                     (loc.loc_bad & LOC_INCONSISTENT ? "(inconsistent)" : "");
             }
-            printf("[PG %u/%u] %lu objects on OSD set %s\n", pool_id, pg_num, stp.second.object_count, osd_set_desc.c_str());
+            printf("[PG %u/%u] %ju objects on OSD set %s\n", pool_id, pg_num, stp.second.object_count, osd_set_desc.c_str());
         }
     }
 }
@@ -468,7 +468,7 @@ void pg_t::calc_object_states(int log_level)
 void pg_t::print_state()
 {
     printf(
-        "[PG %u/%u] is %s%s%s%s%s%s%s%s%s%s%s%s%s%s%s%s%s (%lu objects)\n", pool_id, pg_num,
+        "[PG %u/%u] is %s%s%s%s%s%s%s%s%s%s%s%s%s%s%s%s%s (%ju objects)\n", pool_id, pg_num,
         (state & PG_STARTING) ? "starting" : "",
         (state & PG_OFFLINE) ? "offline" : "",
         (state & PG_PEERING) ? "peering" : "",

src/osd_peering_pg_test.cpp

@@ -49,10 +49,10 @@ int main(int argc, char *argv[])
         pg.peering_state->list_results[osd_num] = r;
     }
     pg.calc_object_states(0);
-    printf("deviation variants=%ld clean=%lu\n", pg.state_dict.size(), pg.clean_count);
+    printf("deviation variants=%jd clean=%ju\n", pg.state_dict.size(), pg.clean_count);
     for (auto it: pg.state_dict)
     {
-        printf("dev: state=%lx\n", it.second.state);
+        printf("dev: state=%jx\n", it.second.state);
     }
     delete pg.peering_state;
     return 0;

src/osd_primary.cpp

@@ -473,7 +473,7 @@ pg_osd_set_state_t* osd_t::add_object_to_set(pg_t & pg, const object_id oid, con
     }
     if (this->log_level >= log_at_level)
     {
-        printf("Marking object %lx:%lx ", oid.inode, oid.stripe);
+        printf("Marking object %jx:%jx ", oid.inode, oid.stripe);
         for (int i = 0, j = 0; i < object_state_bit_count; i++)
         {
             if ((obj_state & object_state_bits[i]) || object_state_bits[i] == 0 && obj_state == 0)
@@ -483,31 +483,31 @@ pg_osd_set_state_t* osd_t::add_object_to_set(pg_t & pg, const object_id oid, con
         }
         if (pg.scheme == POOL_SCHEME_REPLICATED)
         {
-            printf(": %lu copies available", n_copies);
+            printf(": %ju copies available", n_copies);
         }
         else
         {
-            printf(": %lu parts / %lu copies available", n_roles, n_copies);
+            printf(": %ju parts / %ju copies available", n_roles, n_copies);
         }
         if (n_invalid > 0)
         {
-            printf(", %lu invalid", n_invalid);
+            printf(", %ju invalid", n_invalid);
         }
         if (n_outdated > 0)
         {
-            printf(", %lu outdated", n_outdated);
+            printf(", %ju outdated", n_outdated);
         }
         if (n_misplaced > 0)
         {
-            printf(", %lu misplaced", n_misplaced);
+            printf(", %ju misplaced", n_misplaced);
         }
         if (n_corrupted > 0)
         {
-            printf(", %lu corrupted", n_corrupted);
+            printf(", %ju corrupted", n_corrupted);
         }
         if (n_inconsistent > 0)
         {
-            printf(", %lu inconsistent", n_inconsistent);
+            printf(", %ju inconsistent", n_inconsistent);
         }
         printf("\n");
     }